  SEQ CHAPTER \h \r 1 

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

WASHINGTON, D.C. 20460



OFFICE OF

PREVENTION, PESTICIDES AND

TOXIC SUBSTANCES

	PC Code: 084001

DP Barcodes: D290605



 





MEMORANDUM

August 18, 2006





Subject:	EFED response to Public Comments on Dichlorvos Ecological Risk
Assessment



To:	Bob McNally, Branch Chief, Dayton Eckerson, Chemical Review Manager

Special Review Branch

Special Review and Reregistration Division (7508C)

From:	Dana Spatz

  SEQ CHAPTER \h \r 1 Ibrahim Abdel-Saheb

Environmental Risk Branch II 

Environmental Fate and Effects Division (7507C)

Through:	Tom Bailey, Chief, ERB II

Environmental Fate and Effects Division (7507C)



Attached please find EFED’s response to public comments for the
Dichlorvos Ecological Risk Assessment.

Response to Comments Received 8/29/05 from John Kepner, Beyond
Pesticides, 701 E Street, SE #200, Washington, DC 20003

August 29, 2005

Environmental Protection Agency 

Public Information and Records Integrity Branch (7502C)

Office of Pesticide Programs

1200 Pennsylvania Ave., NW

Washington, DC 20460-0001

RE: Dichlorvos (DDVP) Revised Ecological Risk Assessment Docket #
OPP-2002-0302

We appreciate the opportunity to comment on the EPA’s Revised
Ecological Risk Assessment for Dichlorvos (DDVP). These comments are
submitted on behalf of Beyond Pesticides/National Coalition Against the
Misuse of Pesticides (Beyond Pesticides) and our membership. Beyond
Pesticides is an organization committed to pesticide safety and the
adoption of alternative pest management strategies that reduce or
eliminate a dependency on toxic chemicals. To that end, we would like to
use this opportunity to comment on make recommendations to the
Dichlorvos Revised Ecological Risk Assessment. 

Summary

DDVP is an organophosphate (OP) insecticide, which studies show is
highly toxic to birds, aquatic organisms and bees, as well as a probably
human carcinogen. Although the risk it poses to humans is significant,
we will reserve comments on this topic for the IRED and focus our
attention here on the impacts to wildlife and other non-target
organisms. 

DDVP is registered for use in and around homes, schools, parks and other
public spaces, workplaces, greenhouses and more to control flies, gnats,
mosquitoes, chiggers, ticks, cockroaches and insects commonly found on
ornamental plants. It is used in both liquid and granular form. Given
its many uses and formulations, exposure to DDVP, both directly and
through drift, is highly likely to contaminate the natural environment
and expose wildlife to its toxic properties. This likelihood increases
when drift-prone technologies, such as aerial applications, are used.
However, the agency has chosen to ignore certain uses that may expose
wildlife and non-target species, and underestimate the frequency of
other labeled uses.

Even with its underestimations, DDVP exceeds EPA’s Levels of Concern
(LOC) and Risk Quotients (RQs) for various exposure scenarios for both
aquatic and terrestrial organisms. Additionally, EPA’s exposure model
for aquatic organisms does not account for the increase in aquatic
toxicity when DDVP is exposed to UV light. Beyond Pesticides considers
DDVP, like other OPs, outdated technology and not worth the risk its
poses to the environment and human health, given the availability of
alternative products and techniques. We recommend that EPA follow the
lead of the five countries, Angola, Denmark, Fiji, Sweden and the United
Kingdom, that have already banned o the use of DDVP.

Unfortunately, the science behind the DDVP Ecological Risk Assessment is
obscured by the dark shadow cast by EPA’s secret meetings with
DDVP’s manufacturer, as exposed by the Natural Resources Defense
Council lawsuit, NRDC v. Johnson, filed February 17th in federal court
in Washington, DC. While we hope that this assessment has been conducted
honestly and fairly, we will provide further comment on this issue once
a verdict is reached.

Analysis Shows Various Scenarios Exceed EPA LOCs and RQs

EPA’s assessment finds that its level of concern (LOC) is exceeded for
various exposure scenarios including: chronic risk to endangered species
for turf applications (an exposure scenario already underestimated by
EPA (see “EPA Underestimates Worst-Case Scenario Exposures” below)
for birds that consume short grass, tall grass, and broadleaf
plants/small insects (RQ also exceeded); acute risk for bait
formulations for small birds; acute restricted use for bait formulations
for small birds; acute endangered species risk for bait formulations for
small birds; chronic risk for 15 g, 35 g, and 1000 g mammals that
consume short grass, tall grass, and broadleaf plants/small insects in
the turf scenario; endangered species risk for 15 g and 35 g mammals
that consume short grass; acute risk for freshwater invertebrates for
turf scenarios in FL and PA; acute restricted use for freshwater
invertebrates for turf scenarios in FL and PA; and, acute endangered
species risk for freshwater invertebrates for turf scenarios in FL and
PA. 

Comment:

Chronic risk to birds and mammals from the bait formulation is not
currently included, but must be for a complete evaluation.  

EFED Response:

The risk assessment includes chronic toxicity data for birds and mammals
(see pages 31-32) and chronic risk was, in fact, assessed for other uses
besides the bait.  Currently there are no approved methodologies for
determining chronic risk from a granular or bait application of a
pesticide.  Therefore, chronic risk to birds and mammals cannot be
quantified for this particular use.  However, since bait uses are now
limited to premise treatment around barns, feedlots, milk sheds, poultry
houses, and other farm buildings, exposure is less likely and therefore,
adverse chronic effects to birds and mammals are not expected.

Comment:

We would like to reiterate that most of the exposure scenarios that
exceed the agency’s LOC are from turf applications, which we believe
EPA has underestimated, without explanation.

EFED Response:

The use of DDVP on turf is being canceled. 

EPA’s Use Characterization Ignores Uses That Could Expose Wildlife and
Non-Target Species

Comment:

We disagree with EPA’s decision to omit the following exposure
scenarios: liquid DDVP applied around agricultural premises and
structures; liquid and bait applications to manure treatment, garbage
and refuse areas; and, direct treatment to animals. Because deer, ducks,
geese and many other wildlife species have adapted well to the rural
areas surrounding farms, these species often thrive in and around the
very areas being omitted. This scenario cannot be ignored. 

EFED Response:

The risk assessment explained that for these uses, there would be
minimal potential for exposure to terrestrial and aquatic animals based
on the fate properties of dichlorvos, i.e., rapid
degradation/dissipation, and treatment sites being small and localized. 
However, while the risk assessment did not specifically quantitate risk
for or focus on these particular uses, the flying insect scenario
included in the risk assessment covers the risk to birds and mammals for
these uses.  For aquatic species, the turf scenario included in the risk
assessment provides a conservative risk estimate for these secondary
exposure scenarios. 

EPA’s Model Underestimates Impacts on Aquatic Environments

Comment:

In its analysis, EPA uses computer simulation models to estimate
exposure of aquatic organisms, by calculating estimated environmental
concentrations (EECs) in surface water using laboratory data that
describe the rate at which the pesticide breaks down and how it moves
into the environment. However, these models do not always provide
accurate data, due to real world factors not accounted for in the lab.
For example, DDVP becomes 5 to 150 times more toxic to aquatic life when
it is exposed to UV light (U.S. Public Health Service, 1995). Such a
drastic increase in toxicity is extremely important when considering the
impact on aquatic organisms, however it is something that can be easily
overlooked in a computer simulation model. Because of this, we believe
the impact on aquatic species is vastly underestimated.

EFED Response:

Since late 2004, OPP has been using the ECOTOX database on a routine
basis as a search engine to identify open literature studies that may
potentially be used in ecological risk assessments.  However, at the
time the DDVP risk assessment was done, OPP was not routinely performing
ECOTOX-based literature searches, especially when no major data gaps had
been identified in the required Part 158 information submissions.

An ECOTOX search will be scheduled for DDVP during Registration Review
and the risk assessment will be updated as necessary, including
information on possible phototoxicity.

Regarding EPA's use of models, please refer to EFED response to NRDC
General Comment number 2 below.

EPA Underestimates Worst-Case Scenario Exposures

Comment:

For some exposure models, EPA calculates the worst-case scenario
assuming DDVP is applied at the maximum reapplication interval. For
example, granular baits can be reapplied every three to five days, so
EPA assumes the worst-case scenario is 120 applications per year.
However, without explanation, EPA assumes a worst-case scenario of only
four applications for turf applications, despite a 30-day reapplication
interval (or 12 applications per year). This is clearly not the
worst-case scenario, and the risk should be recalculated. 

EFED Response:

The registrant has requested voluntary cancellation of the use on turf.

Comment:

We believe the worst-case scenario is also understated for mosquito
applications, which have no reapplication interval. While the agency
does not believe that mosquito applications are currently used in the
U.S., it must assume an accurate risk based on all legal labeled uses.
Factors such as insect-borne diseases, pesticide resistance and changes
in the regulatory status of other products registered for mosquito use
could prompt the increased use for mosquito control or any other labeled
use.

EFED Response:

The flying insect control use was included in the risk assessment.  This
would include mosquito applications.  While the risk assessment assumed
75 applications per year, the estimated exposure values would change
very little, if at all, if there were even more applications because the
degradation and dissipation of DDVP are so rapid.  DDVP has a foliar
dissipation half life of only 2 hours and a soil metabolism half life of
10 hours.  Thus, there is little or no accumulation from application to
application.  The RQ’s would be essentially the same no matter how
many applications were assumed.  The only difference would be the
probability of contact with a non-target organism.  The more often the
chemical is applied, the greater the chance for exposure.  But, the
overall conclusions of the screening level risk assessment remain the
same.

Impact on Pollinators

Comment:

It should be noted that DDVP is as effective at killing many beneficial
invertebrate species as it is on target pests, so every time DDVP is
applied to where other insects are present, it has the potential to
cause a mass mortality event for beneficial invertebrate species in and
around the sprayed area. This likely includes honey bees (Apis
mellifera) and other pollinators, which continue to decline in the U.S.
with our continued heavy reliance on organophosphorus insecticides.
Extensive data indicate that DDVP poses a very high risk to bees. The
ecosystem service that bees and other pollinators provide to humans is
invaluable and many of our most important food crops require pollination
by bees and other organisms that continue to decline in the face of our
continued heavy reliance on pesticides. 

EFED Response:

Because of the high acute toxicity of DDVP to honey bees, it is expected
that DDVP would be highly acutely toxic to beneficial non-target
insects.  However, a study on the toxicity of residues on foliage to
honey bees showed residues of dichlorvos 4E applied at 0.5 lbs ai/A were
practically non-toxic to honey bees at 3 hours posttreatment.

Conclusion

Our analysis of the Revised Ecological Risk Assessment reveals that EPA
underestimated the risk that DDVP may pose to wildlife. Even using
EPA’s underestimated models, many exposure scenarios still exceed the
agency’s LOCs. Based on wildlife data alone, especially the impact to
pollinators, on which our agricultural system depends, we believe all
remaining uses of DDVP should be cancelled. Regulatory actions must be
swift and not include a potentially harmful phase-out period. We look
forward to analyzing the IRED this fall.

	Sincerely,

	John Kepner 

	Beyond Pesticides 

	701 E Street, SE #200

	Washington, DC 20003	

Response to Comments Received 8/29/05 from Ian S. Chart, AMVAC, 4695
MacArthur Court, Suite 1250, Newport Beach, CA 92660

August 29, 2005

Via E-Mail

Public Information and Records Integrity Branch (PIRIB)

Mail Code 7502C

Office of Pesticide Programs

U.S. Environmental Protection Agency

1200 Pennsylvania Avenue, N.W.

Washington, DC 20460-0001

Re: Dichlorvos (DDVP) Revised Ecological Risk Assessment; Notice of

Availability; Docket ID Number: OPP-2002-0302

Dear Public Information and Records Integrity Branch:

Appended are Amvac Chemical Corporation’s (AMVAC) comments in response
to the United States Environmental Protection Agency’s (EPA) June 30,
2005, Federal Register notice announcing the availability of EPA’s
revised ecological risk assessment for dichlorvos (DDVP). Please let us
know if you have any questions on the appended comments.

Sincerely,

Ian S. Chart

Ian S. Chart

Vice President

Director of Regulatory Affairs

Attachment

09LT060.doc[280.03]

4695 MacArthur Court, Suite 1250, Newport Beach, CA 92660 • (949)
260-1212 • Fax (949) 260-1214

Amvac Chemical Corporation’s

Comments on EFED Dichlorvos (DDVP) Revised Science Chapter

Docket ID Number OPP-2002-0302

August 29, 2005

Comment:

Pages 4 and 5, Executive Summary:

EFED needs to clarify that it’s risk conclusions concerning chronic
risks to birds and small mammals for the turf scenario, and chronic
risks to birds for the flying insect scenario are based on extreme
upper-end estimates of residues on the feed items categories of concern
(short grass, long grass, leafy foliage and small insects, and that
these risk conclusions are based on comparing chronic toxicity endpoints
to acute exposure estimates. This type of comparison represents a
mismatch of toxicity and exposure data, particularly for very
non-persistent pesticides such as dichlorvos.

EFED’s broad, sweeping conclusions concerning risks to aquatic
invertebrates are inaccurate because they are based solely on EFED’s
risk analysis results for freshwater microcrustaceans. There is very
wide diversity among freshwater invertebrates, which these conclusions
fail to recognize. Suggesting all freshwater invertebrates are equally
as sensitive as daphnids to dichlorvos ignores available data on
mollusks that indicate very low toxicity, and no risk concerns for
dichlorvos to these invertebrates.

As detailed below, AMVAC has numerous concerns with EFED’s exposure
modeling and risk analysis procedures for birds, mammals, and aquatic
organisms, and therefore disagrees with EFED’s conclusions concerning
risks to birds, small mammals, and aquatic invertebrates, particularly
EFED’s conclusions concerning chronic risks to birds and small mammals
and EFED’s conclusions concerning acute and chronic risks to
freshwater invertebrates.

EFED Response:

The screening level risk assessment is designed to be conservative and
protective.  In a screening level risk assessment, as described in the
Overview of the Ecological Risk Assessment Process in the Office of
Pesticide Programs (2004) document, which was endorsed by the Fish and
Wildlife Service and the National Marine Fisheries Service, risk
quotients are based on estimated upper bound concentrations and the most
sensitive endpoint is used in calculating risk. We use a standard suite
of surrogate organisms to represent the universe of species in any given
group of organisms, recognizing the variability in toxic response.
However, we believe that just as a number of untested species may be
resistant, there may be other species even more sensitive than those
tested. The Agency is required to protect all species including the
endangered and threatened species.

Comment:

Page 7, Regulatory History: This whole discussion of the regulatory
status of DDVP in other countries should be deleted because it is
irrelevant to the re-registration of DDVP in the U.S., particularly with
respect to EPA’s EFED assessment (all of the points discussed here are
more applicable to human health issues, and therefore, if the regulatory
status of be DDVP is discussed in any of EPA’s documents, it should be
in HED’s assessment, not EFED’s assessment. None of the regulatory
actions in other countries mentioned in this section are based on
environmental or ecological risk concerns.

With respect to carcinogenicity: EPA under its 1999 proposed Guidelines
for Carcinogen Risk Assessment has classified DDVP as having
“suggestive evidence of carcinogenicity but not sufficient to assess
human carcinogenic potential.”

With respect to mutagenicity: There is no evidence of mutagenicity in
vivo.

EFED Response:

The discussion of the status of DDVP in other countries was provided for
informational purposes as background for the problem formulation section
of the risk assessment.  

Comment:

Pages 8 and 9, Use Characterization:

The EFED document is inconsistent and contradictory concerning the
flying insect use pattern because it says both that all available
information indicates that DDVP is not used for outdoor adult mosquito
control, but then it indicates that the 75 applications per year use
scenario evaluated is based on a worst-case estimate of DDVP use to
control adult mosquitoes. Note that elsewhere in the document (see p.
10) it is indicated that ‘For the flying insect (including adult
mosquitoes) use, it does not appear that dichlorvos is being used in
this country’.  The use scenario for dry baits based on 120
applications per year is unrealistic as it assumes pest pressure is
continuous throughout the year. This is also an economically infeasible
and unrealistic use scenario because in reality, if two or three
applications of a product do not control a pest infestation, the user
will try a different product; the user will not continue to use a
non-efficacious product every three days throughout a whole year. This
use scenario represents a misinterpretation of the product label; just
because the label indicates to apply the product at 3 to 5-day intervals
to control a pest infestation does not translate into 120 applications
per year.

EFED Response:

The current labels for DDVP do not state a maximum number of
applications for mosquito and other flying insect control. In the
absence of an upper limit, the Agency modeled 75 applications per year,
which equates to an application approximately every 5 days.  We agree
that this is a conservative estimate considering the pest pressure and
economic reasons given above.  While the risk assessment assumed 75
applications per year, the estimated exposure values would change very
little, if at all, if there were a fewer number of applications because
the degradation and dissipation of DDVP are so rapid.  DDVP has a foliar
dissipation half life of only 2 hours and a soil metabolism half life of
10 hours.  Thus, there is little or no accumulation from application to
application.  The RQ’s would be essentially the same no matter how
many applications were assumed.  The only difference would be the
probability of contact with a non-target organism.  The more often the
chemical is applied, the greater the chance for exposure.  But, the
overall conclusions of the screening level risk assessment remain the
same.

   

We agree that the use scenario for dry baits based on 120 applications
is conservative for the reasons given above, but the product label
allows for use every 3 to 5 days, thus the maximum application could be
120 applications per year.  Again, the estimated exposure values would
change very little, if at all, if there were a fewer number of
applications because the degradation and dissipation of DDVP are so
rapid.  The RQ’s would be essentially the same no matter how many
applications were assumed.  The only difference would be the probability
of contact with a non-target organism.  The more often the chemical is
applied, the greater the chance for exposure.  But, the overall
conclusions of the screening level risk assessment remain the same.

Comment:

Page 10; Figs. 1 and 2, Conceptual Model:

The conceptual model, particularly with respect to liquid formulations
(turf and flying insect control uses) is flawed because it does not
account for or provide any information or perspective on the size of the
area treated for these uses of dichlorvos. In this conceptual model,
dichlorvos is treated and evaluated as if it were a major crop use
agrochemical, which is an inaccurate representation of how dichlorvos is
actually used. The conceptual model for dry baits (Figure 1) also fails
to include any concept of the size of the area treated.

EFED Response:

The conceptual models for the liquid formulation and dry bait are
conservative in the absence of use and usage data that would help to
refine the assessment.  Without verifiable data to the contrary,
widespread use must be assumed.

Comment:

Page 10, Exposure:

The paragraph discussing the flying insect uses of dichlorvos
contradicts EFED’s hypothetical 75 applications per year exposure
scenario (‘For the flying insect (including adult mosquitoes) use, it
does not appear that dichlorvos is being used in this country.’).

EFED Response:

The flying insect (including adult mosquitoes) scenario was included in
the risk assessment because it is a labeled use.

Comment:

Page 11, Receptors of Concern:

EFED should provide specific examples of species that it believes feed
continuously only on, or primarily on (>90% of their diet) short grass,
long/tall grass, and leafy foliage, for both birds, and particularly for
EFED’s theoretical small (15 g) and medium (35 g) mammals. This is
particularly important because EFED has later concluded that birds and
mammals that feed exclusively on these types of feed items are at risk
from DDVP used on turf and for flying insect control.

EFED Response:

The Receptor of Concern section on page 11 is designed to explain the
definitions of herbivore, granivore, and omnivore.  It is not designed
to provide specific examples of species in the problem formulation stage
before ecological risk is assessed.  Additional information will be
provided in the species specific effects assessment for federally listed
threatened and endangered species.

Comment:

Page 11; Figure 2, Aquatic Environment Exposure:

EFED’s hypothesized exposure route of aquatic organisms through
groundwater seepage onto surface water is unrealistic considering the
very rapid degradation of DDVP in soil and DDVP’s instability to
hydrolysis. Elsewhere in this document, EFED concludes that there are no
concerns about DDVP leaching into groundwater because of its rapid
degradation (see p. 22).

EFED Response:

In the Problem Formulation Section, specifically the Aquatic Environment
Exposure Section on page 11, it is necessary to discuss hypothesized
exposure routes that will be further evaluated in the Analysis Section
(i.e. page 22).  Since DDVP is soluble in water, there is potential for
leaching and it is identified as a potential pathway. 

 

Comment:

Page 17, Table 1, Assessment and Measurement Endpoints:

EFED has omitted a key measurement endpoint for evaluating potential
risks to pollinators and beneficial insects, which is the lack of
toxicity to bees after three hours in the residues on foliage study (see
p. 33). Because of this, EFED’s risk conclusions concerning potential
effects on bees and beneficial insects are distorted and biased.

EFED Response:

The screening level risk assessment is designed to be conservative and
protective.  In a screening level risk assessment, as described in the
Overview of the Ecological Risk Assessment Process in the Office of
Pesticide Programs (2004) document, which was endorsed by the Fish and
Wildlife Service and the National Marine Fisheries Service, risk
quotients are based on the most sensitive endpoint. Therefore, the honey
bee acute contact LD50 was used quantitatively in the assessment.  There
is, however, a discussion of the foliar residue bee toxicity study on
page 33 of the risk assessment.

Comment:

Pages 17 and 18, Planned Analyses, Terrestrial Environment:

The planned analyses for potential effects on terrestrial organisms have
no concept of the size of the area treated, which has a significant
impact on risk conclusions. These analyses treat DDVP as a major use
agrochemical, which is inconsistent with how DDVP is actually used.

There are a number of significant assumptions and limitations to
EFED’s conceptual model for terrestrial organism exposure that are not
discussed in this document as uncertainties associated with potential
exposure of non-target organisms, particularly birds and small mammals.
The exposure scenario used by EFED for these analyses assumes a
relatively large treated area (e.g., an agricultural field or sod farm)
in which birds and mammals obtain 100% of their daily diet for both
acute and chronic exposure scenarios; this conceptual model is
inconsistent with the limited outdoor uses of DDVP in which only small
areas are treated. EFED’s analyses also assume that all the feed items
consumed by birds and small mammals contain the maximum estimated
residues based on the modified Hoerger and Kenaga nomogram (Fletcher et
al, 1994).

EFED Response:

Since the current DDVP label fails to provide any limitation to the size
of the area being treated, the risk assessment, in order to be
conservative, must assume relatively large treated areas.  In addition,
the approved screening level risk assessment methodology calls for the
use of the maximum estimated residues based on the modified Hoerger and
Kenaga nomogram (Fletcher et al, 1994) and assumes birds and mammals
obtain 100% of their daily diet for both acute and chronic exposure
scenarios.  The use of maximum estimated residues and the assumption
that birds and mammals consume 100% of their diet in the scenarios gives
a high end estimate of exposure.  This is necessary to be protective of
sensitive bird and mammal species.

Comment:

Page 18, Planned Analyses, Aquatic Environment:

The PRZM-EXAMS modeling scenarios used by EFED for turf are not
appropriate for evaluating the very limited outdoor uses of dichlorvos.
These scenarios assume relatively large treated areas (10 A or 10 ha),
such as a golf course or sod farm, and the models assume 100% of this
area is treated with the pesticide; these types of uses and modeling
scenarios are inconsistent with the use patterns for dichlorvos. The
result is that EECs generated through these modeling scenarios greatly
overestimate aquatic EECs from the very limited outdoor uses of DDVP.

There are a number of significant limitations and assumptions of
EFED’s PRZM-EXAMS modeling scenarios that EFED fails to discuss as
uncertainties associated with the EECs generated through this modeling.
Key assumptions include the conceptual model used in the modeling
scenarios, which are based on a small farm pond completely surrounded by
a 10 ha or 100 A treated field that is 100% treated by the pesticide;
this conceptual model is inconsistent with the very limited outdoor uses
of DDVP in which only small areas are treated. Additional key
assumptions of the PRZM-EXAMS modeling include the receiving water body
(farm pond) directly abuts the treated field (treated area), there is no
water flow through the pond, there is no addition of water to the pond
during runoff events although pesticide is transported into the pond in
runoff water, there is no degradation of the pesticide in the runoff
flow, and the pesticide is instantaneously an homogeneously mixed
throughout the receiving water body. All of these assumptions serve to
maximize and overestimate aquatic pesticide concentrations in the
receiving water body.

Dissipation through volatilization should be incorporated into all of
EFED’s terrestrial and aquatic exposure modeling because this
represents a key route of DDVP dissipation. These use patterns do not
involve wetting-in DDVP, as suggested later in the document by EFED (p.
20).

EFED Response:

EFED agrees that the risk assessment is conservative.  Since the current
DDVP label and use information fail to provide any limitation to the
size of the area being treated, the risk assessment must assume
relatively large treated areas and assume 100% of this area is treated
with the pesticide.  EFED agrees that volatilization can be a major
route of dissipation for DDVP, and improvements to our currently
approved exposure models, with regard to volatilization, are being
considered for future development.  We do note, however, that
volatilization contributes to the rapid foliar dissipation half life for
DDVP, and this was accounted for in the terrestrial risk assessment.

Comment:

Page 22, Aquatic Resource Exposure Assessment:

The turf scenario as described is not consistent with the limited
outdoor uses of dichlorvos, and therefore the EECs generated by this
modeling are unreliable indexes of potential exposure.

EFED Response:

The registrant has requested voluntary cancellation of the use on turf.

Comment:

Page 23, Table 4, PRZM-EXAMS modeling input parameters:

AMVAC notes that EFED has used a DDVP hydrolysis rate for pH 7 for all
its modeling scenarios, including estuarine waters and moderately hard
to hard freshwater where the pH is much closer to 8. However, hydrolysis
of DDVP is much faster at basic pHs (see hydrolysis, p. 21), so using a
hydrolysis rate for pH 7 results in an overestimate of potential
exposure, particularly longer-term exposure, of aquatic organisms in
estuarine-marine and moderately hard to hard freshwater.

AMVAC also notes that EFED has assumed that DDVP is stable to both
aerobic and anaerobic metabolism in aquatic environments (see Table 4,
Appendix C for aerobic and anaerobic aquatic metabolism inputs). This is
erroneous. EFED provides data that DDVP is rapidly metabolized under
anaerobic conditions (DT50 = 6.3 days; see p. 21), so AMVAC wonders why
EFED has not used these data in its PRZM-EXAMS modeling. Likewise, EFED
has assumed that DDVP is stable to aerobic metabolism under aquatic
conditions, apparently because a specific aquatic metabolism study is
not available. However, the aerobic soil metabolism data indicate that
DDVP undergoes very rapid microbial degradation under aerobic
conditions, and therefore AMVAC wonders why EFED did not follow its own
standard procedures and use an estimated aerobic aquatic metabolism rate
of 2 times the aerobic soil metabolism rate for modeling DDVP.

AMVAC also does not believe that EFED’s freshwater farm pond modeling
scenarios are applicable to or representative of estuarine and marine
environments.

EFED Response:

The modeling input parameters were selected in accordance with the
Proposed Interim Guidance for Input Values document, dated April 6,
2000, as noted in the footnote to Table 4.  Because there is currently
no approved estuarine /marine modeling scenario, the freshwater farm
pond was used to estimate EECs for use in calculating estuarine/marine
organism risk.

Comment:

Page 23, Table 5, Aquatic EECs from PRZM-EXAMS modeling:

AMVAC notes that less than 20% of the modeled aquatic loading is due to
spray drift, with >80% of the loading from runoff. Given the very rapid
dissipation of DDVP under actual field conditions (see terrestrial field
dissipation results, p.21), AMVAC believes the high runoff loading
represents an artifact of the PRZM-EXAMS model and the underlying
assumptions. It is very unlikely that under actual use conditions that
any significant quantities of DDVP would be transported to water through
runoff because of this rapid degradation. AMVAC therefore believes that
EFED should re-valuate the risks to aquatic organisms from DDVP based on
spray draft loading only.

EFED Response:

The Agency would gladly review any data generated to confirm lower
runoff concentrations of DDVP.

Comment:

Pages 24 and 25, Terrestrial Organism Exposure Modeling:

As discussed above, AMVAC believes the standard underlying assumptions
for EFED’s terrestrial organism exposure scenarios (both baits and
liquids) are inappropriate for the limited outdoor uses of DDVP that
involve treating only small areas, including the assumptions that birds
and small mammals will obtain 100% of their daily diet from treated
areas, and that all the feed items ingested will contain the maximum
estimated residues, particularly over a longer-term exposure period.

Also, as discussed above, AMVAC believes that EFED’s hypothetical
multiple application scenarios are unrealistic (75 applications per year
for flying insects and 120 applications per year for the dry baits) and
unrepresentative of how DDVP is actually used based on economic and
other considerations.

AMVAC disagrees with EFED’s procedure of extrapolating estimated
residues on broadleaf foliage to insects, and considers this procedure
to not be scientifically valid or supported by available data. There are
no data in the UTAB database (which underlies EFED’s revised nomogram;
Fletcher et al, 1994) on pesticide residues in/on insects. Other, newer
data on pesticide residue in/on insects clearly indicate that pesticide
residues on insects are much lower than those estimated by extrapolating
nomogram-estimated residues for broadleaf foliage to insects (Fischer
and Bowers, 1997; Brewer et al, 1998; EU SETAC meeting, 2005).

The EEC tables (p. 25) should clearly indicate that all the EECs EFED is
using for its risk calculations are initial, peak EECs based on maximum
estimated residues. Based on statistical analyses of the data in the
UTAB database that underlies EFED’s revised nomogram, these maximum
estimated residues represent >99th percentile EECs for short grass and
long grass/tall grass, approximately the 94th percentile EEC for
broadleaf foliage, and approximately the 98th percentile EEC for seeds
and fruits, so there is a correspondingly very low probability that
birds or small mammals would actually ingest feed items containing
residues at these estimated concentrations, particularly on a
longer-term (chronic) basis. The distributions of residues for each of
the four feed item categories that EFED evaluates (short grass, long
grass/tall grass, broadleaf foliage, seeds and fruits) are all
log-normal, and AMVAC believes that EFED should conduct a more through
analysis by evaluating potential EECs for median estimated residues
rather than just extreme upper bound estimated residues; this type of
analysis would put potential risks into a more appropriate perspective.

AMVAC also disagrees with EFED’s approach to chronic risk assessment
for birds and small mammals in which an initial, acute EEC is compared
to a NOEC from a longer-term study, particularly for very non-persistent
products such as dichlorvos. EFED attempts to justify this approach by
citing Bennett and Ganio, 1991 (p. 32). However, other work by these
same authors has indicated that NOECs for short-term exposures are
higher than those for longer-term exposures, and for OPs, many of the
short-term effects that were noted on birds reflect nutritional effects
(e.g., reduced feeding) rather than direct reproductive toxicity from
Ops (Bennett and Bennett, 1990, Bennett et al, 1990). The effects listed
by EFED from avian reproduction testing (p. 31) are not effects that
occur with a single exposure, particularly at the NOEC and LOEC
concentrations, so comparing the NOEL to the instantaneous EEC is
inappropriate. If EFED is concerned about possible effect occurring in a
relatively short time period, then EFED should compare the NOEC and LOEC
to a time-weighted exposure for that short time period.  Alternatively,
EFED should compare the estimated residues (both maximum and median
EECs) on a daily basis to the NOEC and LOEC to estimate the time period
that estimated residues might exceed the NOEC and LOEC. Given the very
rapid dissipation of DDVP from foliage (DT50 << 1 day, p. 24), it is
clear that chronic exposure of birds and small mammals does not occur
from DDVP.

The LD50/square foot calculation that EFED relies on for evaluating
potential exposure and risk from the bait formulation has never been
validated. Available evidence from avian field studies with granular
products clearly indicates that other factors, rather than the amount of
pesticide per unit area, are more significant determinants of the
potential risk associated with granular products.

EFED has omitted some key information from its assessment procedure for
the dry baits. This information includes the size of the granule and the
amount of a.i. on each granule.

EFED Response:

The DDVP ecological risk assessment is a screening level risk assessment
designed to be conservative and protective.  It was conducted according
to the guidance described in the Overview of the Ecological Risk
Assessment Process in the Office of Pesticide Programs (2004) document,
which was endorsed by the Fish and Wildlife Service and the National
Marine Fisheries Service.  The use of maximum estimated residues and the
assumption that birds and mammals consume 100% of their diet in the
scenarios gives a high end estimate of exposure.  This is necessary to
be protective of sensitive bird and mammal species.  As discussed above,
the selection of number of applications is appropriate given the wording
on the current DDVP labels.

Comment:

Page 27, Table 11:

Clarify why one Study Classification is considered ‘core’ and the
other two ‘supplemental’ when all the data are from the same source;
AMVAC notes that the most commonly tested daphnid species is Daphnia
magna, not Daphnia pulex.

EFED should modify the sentence beneath Table 11 to clarify that based
on the available data, DDVP is classified as very highly toxic to
crustaceans, rather than broadly extrapolating these results to all
freshwater invertebrates, particularly because data on mollusks (p. 29)
indicates that DDVP is of very low toxicity (classified as slightly
toxic) to this major group of aquatic invertebrates.

EFED Response:

Studies are classified core, supplemental or invalid.  Core studies
conform to EPA guidance, while supplemental studies have minor
deviations but are still scientifically sound. Several studies may be
from the same source and have different classifications.  EFED believes
the sentence beneath Table 11 is accurate as approved risk assessment
methods support extrapolation from surrogate species.

 

Comment:

Page 29, Acute Toxicity to Estuarine and Marine Invertebrates:

The 1st sentence beneath table 15 categorizes dichlorvos as very highly
toxic to estuarine/marine animals on an acute basis. To be accurate,
dichlorvos is classified as very highly toxic to slightly toxic to
estuarine/marine invertebrates, depending on the species tested and
group represented; it is not appropriate or accurate to simply describe
dichlorvos as very highly toxic to estuarine invertebrates.

EFED Response:

The approved risk assessment methodology uses the most sensitive
endpoint.  Therefore, DDVP is considered very highly toxic based on the
mysid LC50 of 19.1 ppb and 18.7 ppb for formulated product. 

Comment:

Page 31, Subacute Dietary Toxicity:

The 1st sentence beneath table 19 categorizes dichlorvos as being
moderately toxic to avian species on a subacute dietary basis. To be
accurate, dichlorvos is classified as moderately toxic to practically
non-toxic to young birds, depending on the species; it is not
appropriate or accurate to simply describe dichlorvos as moderately
toxic birds on a subacute dietary basis.

EFED Response:

The approved risk assessment methodology uses the most sensitive
endpoint.  Therefore, DDVP is considered moderately toxic based on the
pheasant LC50 of 568 ppm. 

Comment:

Page 33, Toxicity to Insects:

The results of the toxicity of residues on foliage to bees study should
be included as a measurement endpoint for evaluating and characterizing
potential risks of dichlorvos to pollinators and beneficial insects.

EFED Response:

This study was discussed in the risk assessment.  The approved risk
assessment methodology uses the most sensitive endpoint.  Therefore,
DDVP is considered highly toxic based on the bee LD50 of 0.495 (g/bee.

Comment:

Page 34, Probit Slope Response Relationship:

EFED should provide a copy of the referenced model to registrants so
that they can independently evaluate EFED’s calculations and
conclusions based on this model. EFED should also provide examples of
these calculations to make this process transparent, as well as some
insight as to how EFED interprets these calculations.

EFED Response:

The Individual Effect Chance Model used in probit slope analysis is a
simplistic Excel spreadsheet tool.  It is based on a basic tenet of
toxicology, a dose response relationship for toxicants, and uses a mean
estimate of slope and assumes a probit slope response relationship. In
the analysis on pages 46 to 47, the raw data, assumptions, and results
are very clearly stated and transparent with clear interpretations of
the results.

Comment:

Pages 35 and 36, Incidents:

The reference to the aquatic incident in Tennessee should be deleted
unless additional information is provided to better define the incident,
particularly as EFED indicates no residue analysis was conducted. EFED
has not presented any information clearly linking this reported incident
to DDVP, much less to registered uses of DDVP (i.e., if the incident is
related to DDVP, it may very well be a misuse).

The discussion in the last two paragraphs on p. 35 and the first two
paragraphs on p. 36 is irrelevant to dichlorvos and dichlorvos use
patterns. These concern agricultural uses of pesticides, not indoor uses
or uses around buildings such as those for which dichlorvos is
registered. This paragraph should be deleted. The likelihood of
detecting incidents from uses such as those for which dichlorvos is
registered is much higher than for agricultural uses because these uses
are in confined areas without lots of cover where people are more likely
to notice an incident, the uses are in areas where people are commonly
present and, as exemplified in the incidents discussed above, people are
more likely to report these because the incidents are more likely to
involve animals valuable to the people. The referenced Balcomb (1986)
report is hardly the only report on carcass removal rates from
agricultural fields; the numerous avian field studies conducted on
pesticides routinely measure carcass removal rates, and these studies
indicate that carcass removal rates are highly variable, even for fields
in close proximity; in some cases carcasses are removed rapidly as
reported by Balcomb, and in others carcasses persist much longer.
Therefore, it is inappropriate and inaccurate for EFED to merely cite
and extrapolate the results of one study conducted on a couple of fields
in Maryland as support for EFED’s bias concerning wildlife incident
reports.

Interestingly, but not mentioned by EFED, there are several chemicals
that have extensive avian or aquatic incident histories which suggests
that incidents are not nearly as unseen or unreported as EFED claims;
reported incidents involve as few as a single small bird, or a few small
fish.

AMVAC disagrees with EFED’s conclusion (second paragraph, p. 36) that
the reporting of a few dead birds associated with the use of a chemical
is believed to provide evidence that substantial effects may be
occurring. There is simply no data or evidence to support this extreme
conclusion; rather this conclusion very much appears to be grasping at
straws to try to support hypothetical risk assessment conclusions.

EFED Response:

The reference to the aquatic incident in Tennessee is appropriately
reported as being related to DDVP use even though residue analysis was
not conducted.  

Because the U.S. does not have a formal, mandated incident reporting
system, this suggests the reported incidents are a small fraction of
actual incidents.  The avian and mammalian incidences reported are
detailed and substantiated.

Comment:

Page 37, Risk Characterization, Freshwater Invertebrates:

As detailed above, AMVAC has numerous concerns with the aquatic modeling
used to estimate aquatic EECs for the turf uses of DDVP, and for these
reasons believes the EECs used by EFED to calculate risk quotients
greatly overestimate potential exposure of aquatic organisms to DDVP.
AMVAC therefore disagrees with EFED’s risk conclusions concerning
freshwater invertebrates, and particularly EFED’s extrapolation of its
risk conclusions based on daphnids to all other freshwater taxonomic
groups.

EFED Response:

The registrant has requested voluntary cancellation of the use on turf.

Comment:

Pages 38 and 39, Risk Characterization, Estuarine and Marine Organisms:

Although EFED’s risk calculations do not indicate significant acute or
chronic risks to estuarine or marine organism, AMVAC has a number of
concerns, as detailed above, with the applicability of EFED’s farm
pond turf modeling scenarios to estimate potential exposure of estuarine
and marine organisms.

EFED Response:

Because there is currently no approved estuarine /marine modeling
scenario, the freshwater farm pond was used to estimate EECs for use in
calculating estuarine/marine organism risk.

Comment:

Pages 39 and 40, Risk Characterization, Birds:

As detailed above, AMVAC has numerous concerns with EFED’s exposure
and risk methodologies for evaluating potential risks of DDVP to birds,
particularly EFED’s chronic risk  methodology. For all of these
reasons, AMVAC disagrees with EFED’s risk conclusions for birds,
particularly EFED’s chronic risk conclusions.

Tables 28-30 should provide the EECs that EFED is using for its risk
quotient calculations.  These tables should clearly indicate that all
the EECs EFED is using for its risk calculations are initial, peak EECs
based only on maximum estimated residues.

EFED Response:

The approved screening level risk assessment methodology uses the
maximum estimated residues for risk calculations.  See Appendix D for
the endpoints, EECs, doses, and calculated RQs for each terrestrial
scenario modeled.

Comment:

Pages 41 and 42, Risk Characterization, Mammals:

As detailed above, AMVAC has numerous concerns with EFED’s exposure
and risk methodologies for evaluating potential risks of DDVP to
mammals, particularly EFED’s chronic risk methodology.

Interestingly, if chronic dietary risks to small mammals are calculated
using the appropriate procedure (comparing the dietary reproduction NOEC
of 20 ppm to estimated dietary concentrations on feed items), there are
no chronic risks to small mammals for any feed item category, even using
EFED’s extreme EEC estimates (initial maximum residues). This is the
procedure that EFED used for its chronic avian assessment, and the same
procedure should be applicable to small mammals.

As mentioned above, AMVAC requests that EFED provide specific examples
of small mammals that it believes feed exclusively on short grass, long
grass/tall grass, and broadleaf foliage because these are the feed items
that EFED has concluded pose acute and chronic risks to small mammals.
As mentioned above, AMVAC disagrees with EFED’s procedure of lumping
insecttype feed items in with broadleaf foliage because all the
available data indicate that extrapolation is erroneous.

In addition to the problems with EFED’s mammalian risk assessment
procedures discussed above, EFED’s estimated LC50 procedure for
estimating a toxicity endpoint for small mammals (or, conversely,
estimating a daily dose from residue concentrations on feed items) for
calculating risk quotients using the T-rex model is flawed. The
estimated LC50 is driven by the estimated daily feed consumption for
each of the size mammals that EFED evaluates (15 g, 35 g, and 1000 g).
EFED estimates that the 15 g mammals ingests 95% of its body weight in
feed each day, resulting in an estimated LC50 value used as the toxicity
endpoint that is very nearly the same as the mammalian LD50 value. For
the 35 g animal, EFED estimates that animals of this size ingest 66% of
their body weight in feed daily, resulting in an estimated LC50 toxicity
endpoint that is only about 30% higher than the acute LD50 value. Actual
available for small mammals of these sizes (shrews, mice, voles) do not
support EFED estimated high daily feed ingestion rates (EPA Wildlife
Exposure Factors Handbook, 1993) An additional flaw in EFED’s
mammalian risk assessment methodology is that EFED assumes that residues
in/on feed items have the same toxicity as a bolus oral dose of a
pesticide. However, actual data again indicate that this assumption is
erroneous, as acute oral toxicity values are known to be poor predictors
of dietary toxicity (McCann et al, 1981).

Intuitively, EFED risk conclusions for small mammals do not make sense
when compared to EFED’s risk calculations for birds. The toxicity data
on DDVP indicates that DDVP is more toxic to birds than mammals by
approximately four to eight-fold on both and acute and chronic basis,
yet EFED has calculated higher risk quotients for small mammals than for
birds using the same EECs. This incongruous result indicates that
EFED’s risk methodology for evaluating potential risks to small
mammals is flawed.

Tables 31 though 33 indicate a mammalian reproduction NOEC of 5 ppm; the
correct value is 20 ppm. It is unclear whether this error is simply a
typographical error in the legends for these tables, or whether this
erroneous value was used in EFED chronic mammalian risk calculations.

For all of these reasons, AMVAC disagrees with EFED’s risk conclusions
for small mammals, particularly EFED’s chronic risk conclusions.

EFED Response:

The DDVP ecological risk assessment is a screening level assessment. 
The screening level risk assessment, as described in the Overview of the
Ecological Risk Assessment Process in the Office of Pesticide Programs
(2004) document, which was endorsed by the Fish and Wildlife Service and
the National Marine Fisheries Service, is designed to be conservative
and protective.  Please see Appendix D of the risk assessment for
clarification of appropriate mammalian equations. The tables 31 to 33
legends are indeed a simple typographical error.  Please see Appendix D
for clarification that the Rat NOEC of 20 ppm was used in all risk
calculations.

Comment:

Page 42, Risk Characterization, Non-Target Terrestrial Invertebrates:

This risk characterization is incomplete because it does not consider
the results of the toxicity of residues on foliage study in evaluating
potential risks to pollinators and beneficial insects.

EFED Response:

This study was discussed in the risk assessment.  The approved risk
assessment methodology uses the most sensitive endpoint.  Therefore,
DDVP is considered highly toxic based on the bee LD50 of 0.495 (g/bee.

Comment:

Page 44, Risk Description, Summary of Major Conclusions for Aquatic
Animals:

As detailed above, AMVAC has numerous concerns with EFED’s exposure
modeling and risk analysis procedures for aquatic organisms, and
therefore disagrees with EFED’s conclusions concerning risks to
freshwater invertebrates.

EFED Response:

EFED’s risk conclusions for freshwater invertebrates were derived
based on SAP approved models and methodologies.  The exposure modeling
and risk analysis procedures have also been endorsed by the Fish and
Wildlife Service and the National Marine Fisheries Service.

Comment:

Page 44, Risk Description, Summary of Major Conclusions for Terrestrial
Animals:

As detailed above, AMVAC has numerous concerns with EFED’s exposure
modeling and risk analysis procedures for birds and mammals, and
therefore disagrees with EFED’s conclusions concerning risks to birds
and small mammals, particularly EFED’s conclusions concerning chronic
risks.

EFED Response:

EFED’s risk conclusions for birds and mammals were derived based on
SAP approved models and methodologies.  The exposure modeling and risk
analysis procedures have also been endorsed by the Fish and Wildlife
Service and the National Marine Fisheries Service.

Comment:

Pages 45-50, Threatened and Endangered Species Concerns:

As detailed above, AMVAC has numerous concerns with EFED’s exposure
modeling and risk analysis procedures for birds, small mammals, and
aquatic organisms, and therefore disagrees with EFED’s conclusions
concerning risks to birds, small mammals, and aquatic organisms (pp. 45
and 46). These particularly include EFED’s conclusions concerning
chronic risks birds, acute and chronic risk to small mammals, EFED’s
conclusions concerning risks of the dry bait to birds, and EFED’s
conclusions concerning acute and chronic risks to freshwater
invertebrates.  A clear, detailed discussion of the use patterns that
suggest concerns is missing from the endangered species section; much of
the concern for potential effects on endangered species effects would
likely be resolved with a more detailed and focused consideration of
dichlorvos uses.

The table (Table 35) on p. 46 does not provide any useful information.
If EFED believes that specific endangered species are potentially at
risk from dichlorvos, then it needs to specify what those species are
and where they are found, and what their habitat and diets are that
result in them being in close proximity to where dichlorvos is likely to
be used. In contrast, this table suggests that all these threatened and
endangered species in all of these taxonomic groups are potentially at
risk from the very limited outdoor uses of dichlorvos. This is clearly
erroneous, and even contradicts EFED’s own conclusions for most
taxonomic groups.

EFED needs to provide some interpretation of its probit slope response
calculations (p. 47); as presented, these calculations are meaningless.
As presented, EFED’s calculations suggest that dichlorvos uses may
cause widespread mortality among birds, small mammals, and aquatic
invertebrates. However, the numerous concerns with EFED’s methodology
and use pattern considerations do not support such a conclusion.

EFED’s conclusions concerning potential indirect effects to
endangered/threatened and its discussion of critical habitat are based
on the presumption that EFED’s assessments and conclusions about
potential effects to freshwater invertebrates, birds and small mammals
are valid. As discussed above, for numerous reason, AMVAC des not
believe that EFED’s risk procedures and risk conclusions for these
groups of organisms are valid.

EFED Response:

EFED agrees that the screening level risk assessment is conservative and
protective and that concern for endangered species could be resolved
with a more detailed and focused consideration of DDVP uses.  We look
forward to receiving additional use and usage data in order to conduct a
refined species specific assessment for federally listed threatened and
endangered species.Response to Comments Received 8/29/05 from Aaron
Colangelo, Natural Resources Defense Council, 1200 New York Ave NW,
Suite 400, Washington DC 20005

"Colangelo, Aaron" <acolangelo@nrdc.org>

08/29/2005 05:36 PM To

Group OPP-Docket@EPA

cc

Dayton Eckerson/DC/USEPA/US@EPA

Bcc

Subject

OPP-2002-0302

Attached are NRDC’s comments on EPA’s Revised Ecological Risk
Assessment

for DDVP, in response to the notice of availability published at 70 Fed.

Reg. 37844 (June 30, 2005).

Thank you,

Aaron Colangelo



August 29, 2005

Public Information and Records Integrity Branch (PIRIB)

Office of Pesticide Programs, 7502C

Environmental Protection Agency

1200 Pennsylvania Ave, NW

Washington, DC 20460-0001

  HYPERLINK "mailto:opp-docket@epa.gov"  opp-docket@epa.gov 

Attn: Docket ID Number OPP-2002-0302

Re: Dichlorvos (DDVP) Revised Ecological Risk Assessment

These comments are submitted in response to EPA’s request for comment
on the Dichlorvos (DDVP) Revised Ecological Risk Assessment and risk
mitigation measures, published at 70 Fed. Reg. 37844 (June 20, 2005),
and to assist the Agency in its deliberations on final decisions in the
DDVP Special Review, reregistration, and tolerance reassessment
processes. NRDC requests that EPA fully consider these comments before
it makes those decisions and respond to them at or before the time it
issues those decisions.

These comments are organized into two categories. The first section
contains general comments directed at process, policy and science issues
in connection with the DDVP review processes. The second section
contains specific comments indexed to specific pages in EPA’s Revised
Ecological Risk Assessment for DDVP, on which EPA is seeking public
comment.

GENERAL COMMENTS

Comment:

1. The DDVP reregistration and tolerance reassessment processes have
been carried out against the background of the longest ongoing Special
Review in the history of the Agency. The story of Agency attempts to
regulate DDVP based on safety concerns begins in 1982: “EPA first
expressed concern with the safety of DDVP in 1982.  Following its review
of a carcinogenicity study conducted by the National Cancer Institute,
EPA proposed to require a cancer warning on DDVP products in the 1987
DDVP Registration Standard. EPA deferred imposition of the warning and
initiated the ongoing Special Review of pesticide products containing
DDVP, based in part on EPA’s expectation that DDVP would be classified
as a B2 carcinogen.” (Amvac letter to EPA, Nov. 20, 1998).

DDVP was first placed in Special Review in 1988. The ensuing long
history of EPA’s abortive attempts to regulate DDVP through the
Special Review process is replete with unexplained delays. The first
stage of the Special Review began in 1988 and dragged on for an
extraordinary seven years before a preliminary decision was issued.
(PD-2/3, Sept. 28, 1995). The PD 2/3 proposed cancellation of most uses
of DDVP. Five years after the publication of the PD 2/3, the proposed
cancellations had not been implemented, but EPA predicted that the final
decision, in the form of a PD-4 would be issued later that year or in
2001. (Status of Chemicals in Special Review, March 2000, at 21). The
Agency missed this deadline and has continued to fail to implement the
preliminary determination of the Special Review. It has never explained
its inaction. After 17 years the Special Review remains uncompleted.
During these 17 years OPP staff and upper level managers have met with
the registrants on numerous occasions, sometimes secretly and often in
response to aggressive letters from Amvac.

In the midst of the Special Review, EPA was compelled by amendments to
FIFRA and the FFDCA to undertake both the reregistration of DDVP
registrations and the reassessment of the associated tolerances
(together, the “RED process”). The simultaneous conduct of these
three review processes was apparently treated by EPA as a single
process. Thus, EPA’s frequent meetings with Amvac would often involve
issues common to all three processes, although that fact was rarely, if
ever, acknowledged in the reports of those meetings. The unusual access
granted Amvac – both in terms of the high level of the EPA
representatives in attendance and the frequency of the meetings –
eventually became a source of concern and embarrassment to EPA,
culminating in an extraordinary statement by the Agency expressing
regret over the improper access it had granted Amvac. The Agency
admitted that “EPA has provided Amvac with more opportunity to
participate in the preliminary risk assessment process prior to public
involvement than intended by the standard process developed at TRAC. In
retrospect, allowing Amvac to provide its arguments on the preliminary
risk assessment prior to public involvement may have been a mistake.”
(EPA reply to Amvac’s letter of May 5, 1999, undated, at 1).

The frequent secret meetings with Amvac in conjunction with Amvac’s
incessant requests for delay appear to have produced the 17 year (and
still counting) Special Review. The unreasonable and unexplained delays
in the Special Review process have been mirrored by similar delays in
the reregistration and tolerance reassessment processes. Although the
Agency’s preliminary risk assessment (PRA) for DDVP was sufficiently
refined to be privately shared with Amvac for “error correction” on
December 11, 1998, the “corrected” version of this PRA was not made
available to the public until April 15, 2000. The Special Review and RED
processes have completely stalled since the 2000 public release of the
PRA. The public has been given no explanation for these extraordinary
and apparently unprecedented delays.

Public health and the environment have both been compromised by EPA’s
unreasonable delay in completing the Special Review and the RED process.
The Agency owes the public a clear and specific explanation for the
extraordinary delays in the 17 year Special Review and 7 year RED
process for DDVP. These delays represent a complete failure on the part
of the Agency to carry out its duty to protect the public health and the
environment. The Agency must immediately conclude the Special Review,
reregistration, and tolerance reassessment processes, issue a notice of
intent to cancel all DDVP registrations, and initiate revocation actions
for all tolerances for DDVP residues in food and feed.

EFED Response:

This comment relates to the special review and reregistration of DDVP
generally rather than the ecological risk assessment.  The Agency will
address this comment in its response to comments on the IRED.

Comment:

2. In many instances, the Agency relies on models rather than empirical
data because of data deficiencies that are the result of the Agency’s
failure to collect adequate data during the Special Review and RED
processes. There is no evidence in the record that any of these models
was validated by means of empirical studies designed to test their
accuracy. Neither is there any evidence that the Agency has tried to use
monitoring data to confirm the accuracy of its models. The advice of
scientific peer review groups concerning these models can be useful, but
it is not a substitute for true validation by means of empirical data
derived from studies designed to test the accuracy of the model
predictions. While some of these models can be used as screening models,
their design and input shortcomings render them far too insensitive to
reliably demonstrate the absence of risk or exposure. Thus, they are far
more likely to produce false negative results than false positives.
Accordingly, these models should never be used in support of a decision
to continue a registration or tolerance. They can be used as evidence of
possible risk or exposure, and thereby create a presumption of risk or
exposure that the registrants must rebut. In light of the powerful
authorities available to the Agency to compel production of the needed
data – and the clear expectation of Congress, as expressed in Section
4 of FIFRA and Section 408 of the FFDCA, that EPA obtain any data needed
to demonstrate the safety of pesticides regulated under those statutes
– there is no excuse for EPA’s failure to obtain any needed data
prior to the completion of any review process.

EFED Response:

Both EPA and stakeholders in the United States and internationally use
PRZM and EXAMS to estimate pesticide concentrations in water as a result
of agricultural use. These models are generally recognized as suitable
tools for assessment of exposure in aquatic environments.  As documented
in a number of validation efforts (see listing below) and in FIFRA
Science Advisory Panel external peer reviews held in July 1998 and May
1999; 

  HYPERLINK
"http://www.epa.gov/scipoly/sap/meetings/1998/july/1part4.pdf" 
http://www.epa.gov/scipoly/sap/meetings/1998/july/1part4.pdf 

  HYPERLINK
"http://www.epa.gov/scipoly/sap/meetings/1999/may/pca_sap.pdf" 
http://www.epa.gov/scipoly/sap/meetings/1999/may/pca_sap.pdf 

the models generate appropriately conservative estimates of pesticide
concentrations in water that do not significantly underestimate
pesticide concentrations that could actually be found in water bodies
vulnerable to high pesticide runoff, while not generating unrealistic
estimates that are so high that pesticides that do not pose an actual
concern or true risk pass the screening evaluation.

OPP typically relies on field monitoring data as well as mathematical
models to generate exposure estimates for both drinking water and
aquatic exposure assessments and for water quality assessments. 
Monitoring and modeling are both important tools for assessing pesticide
concentrations in water and can provide different types of information.
Monitoring tells us what is happening under current use practices and
under typical conditions. Although monitoring data can provide a direct
estimate of the concentration of a pesticide in water, it does not
always provide a reliable estimate of acute exposure because sampling
may not occur where the highest concentrations of a pesticide are found,
and/or the sampling may not occur when the pesticides are being used.

OPP typically relies on mathematical models to generate the exposure
estimates for drinking water and aquatic exposure assessments and water
quality assessments to better understand the extent to which monitoring
data may underestimate exposures, both spatially and temporally. These
models calculate estimated environmental concentrations (EECs) using
laboratory data that describe how fast the pesticide breaks down to
other chemicals and how it moves in the environment. Although computer
modeling provides an indirect estimate of pesticide concentrations,
these concentrations can be estimated continuously over long periods of
time, and for places that are of most interest for any particular
pesticide. Modeling is a useful tool for characterizing vulnerable
sites, and can be used to estimate peak concentrations from infrequent,
large storms. These concentrations are compared to toxicity data for the
pesticide to determine the risk posed in both drinking water and aquatic
systems. Computer modeling can also be used to determine how different
mitigation practices affect the amount of the pesticide which can run
off into water.

	In estimating pesticide concentrations in water resources, OPP uses a
tiered screening approach involving a combination of models and
monitoring data. The intent of the screening approach is to estimate
pesticide concentration in water from sites that are highly vulnerable
to runoff or leaching so that the program can be confident that any
pesticide that passes the screening tiers poses a low possibility of
harming human health, wildlife, or the environment. Failing a tier,
however, does not necessarily mean the chemical is likely to cause
health or environmental problems, but rather that there is a need to
move to a higher tier and conduct a more refined assessment. This tiered
modeling approach is designed to provide a thorough analysis of each
pesticide, while at the same time focus OPP's efforts on those
pesticides that pose the greatest potential risk.

Monitoring data and modeling estimates provide different types of
information that are used together and in a complimentary way to
characterize pesticide exposure.  It would not be feasible to rely
solely on monitoring data; for the reasons explained above, not to
mention the fact that it would be very inefficient, if not impossible,
to capture all relevant scenarios, site and meteorological conditions,
soil types, etc., in a monitoring program.

References on PRZM and EXAMS validation efforts

Allen, B. Walter, jr, M. Craig Barber, Sandra Bird, Lawrence, A. Burns,
J. Mark Cheplick, Dudley R. Hartel, Charles A. Kitner, Foster L. Mayer,
Luis Suarez, and Steven E. Wooten. 1991. PIRANHA: Pesticide and
Industrial Chemical Risk Analysis and Hazard Assessment, Version 2.0. p
21-35. Environmental Research Laboratory, Office of Research and
Development, U. S. Environmental Protection Agency, Athens, GA.

Durborow, T. E., N. L Barnes, S. Z. Cohen, G. L. Horst, A. E. Smith.
2000. Calibration and Validation of Runoff and Leaching Models for Turf
Pesticides, And Comparison with Modeling Results. Chapter 12 in Fate and
Management of Turfgrass Chemicals. ACS Symposium Series 743. American
Chemical Society.

Jones, Russell L., and Mark H. Russell (eds.). 2001. FIFRA Environmental
Modeling Validation Task Force: Final Report. American Crop Protection
Association.Washington,DC.   HYPERLINK
"http://femvtf.com/femvtf/Files/FEMVTFbody.pdf" 
http://femvtf.com/femvtf/Files/FEMVTFbody.pdf 

Kolset, Knut and Anders Heiberg. 1988. Evaluation of the ‘Fugacity
(FEQUM) and the ‘EXAMS’ Chemical fate and Transport Models: A Case
Study on the Pollution of the Norrsundet Bay (Sweden). Wat. Sci. Tech.
20:1-12.

Melancon, Susan. M., James M. Pollard, and Stephan C. Herd. 1986.
Evaluation of SESOIL, PRZM and PESTAN in a Laboratory Column Leaching
Experiment. Environmental Toxicology And Chemistry 5:865-878.

Mueller, Thomas C., Ronald E. Jones, Parshall B. Bush, and Phillip
Banks. 1992. Environmental Toxicology and Chemistry 11:427-436.

Office of Pesticide Programs. 2002. Revised OP Cumulative Assessment:
Appendix E1: Comparisons of Estimated Regional OP Distributions with
Occurrences in Ambient Water from the USGS NAWQA Program.   HYPERLINK
"http://www.epa.gov/pesticides/cumulative/rra-op/III_E_WaterAppendix/III
_E_1.pdf" 
http://www.epa.gov/pesticides/cumulative/rra-op/III_E_WaterAppendix/III_
E_1.pdf 

Parrish, Rudolph S., Charles N. Smith, and J. Mark Cheplick. 1989.
Testing Pesticide Leaching Models: An Agricultural Field Study.
Environmental Research Laboratory, Athens, GA. 

Parrish, Rudolph S., Charles N. Smith, and Fred K. Fong. 1992. Tests of
the Pesticide Root Zone Model and the Aggregate Model for Transport and
Transformation of Aldicarb, Metolachlor, and Bromide. Journal of
Environmental Quality 21(4):685-697.

Smith, C. N., R. S. Parrish, R. F. Carsel, A. S. Donigian, jr, and J. M.
Cheplick. 1988. Validation Status of the Pesticide Leaching   SEQ
CHAPTER \h \r 1 and Groundwater Transport Model. Environmental Research
Laboratory, Office of Research and Development. U. S. Environmental
Protection Agency. Athens, GA.

Walker, Allan, R. Calvet, A. A. M. Del RE, W. Pestemer, and J. M.
Hollis. 1995. Evaluation and Improvement of Mathematical Models of
Pesticide Mobility in Soils and Assessment of Their Potential to Predict
Contamination of Water Systems. Blackwell Wissenschafts-Verlag GmbH,

Comment:

3. The Agency uses many assumptions both in its models and in connection
with simple calculations. Virtually none of these assumptions appears to
have been validated with data. Indeed, many of the assumptions appear to
be based on anecdotal information that has not even been confirmed by
independent means. Some of the sources of information are industry
groups and other parties with a financial interest in the outcome of the
DDVP decisions. Information from such sources should be subject to
careful scrutiny and rigorous validation. This does not appear to have
been done.

EFED Response:

It is not clear which assumptions are in question.  The Agency gathers
the best available data, including use and usage, toxicity, and
fate/transport information in order to conduct an ecological risk
assessment.  Before any piece of information is included in the risk
assessment, it undergoes careful review and scrutiny according to Agency
guidelines, policies and procedures.  The Agency also characterizes
uncertainties when data are missing and/or could not be fully validated.


Comment:

4. Despite 17 years in Special Review and at least 7 years in the RED
process, EPA failed to require submission of data to replace missing or
inadequate data. In many instances EPA even acknowledged the weakness of
studies, but provided no explanation for its failure to call in the
appropriate data. Rather than calling in needed data, the Agency
frequently substituted unvalidated models that, in turn, relied on
unsubstantiated assumptions. These models are not an adequate substitute
for empirical data and they do not relieve the Agency of its duty to
obtain any data necessary to make the safety findings required by FIFRA
and FFDCA. EPA should have required the appropriate studies to be
submitted in the early stages of the Special Review and RED processes.
In the absence of an adequate database for DDVP, EPA cannot conclude
that DDVP will not cause unreasonable adverse effects on the environment
or that the tolerances for DDVP are safe.

Nor can EPA use the inadequacy of the DDVP database as an excuse for
further delaying regulatory action against DDVP. It is as much, if not
more, the registrant’s responsibility to maintain adequate scientific
support for its registrations and associated tolerances. The
responsibility of demonstrating the safety of registrations and
tolerances is always on the registrant. Accordingly, all DDVP
registrations should be cancelled and all tolerances for DDVP should be
revoked.

EFED Response:

This comment relates to the special review and reregistration of DDVP
generally rather than the ecological risk assessment.  The Agency will
address this comment in its response to comments on the IRED.  The
adequacy of the aquatic exposure models employed in the ecological risk
assessment is addressed in response to comment number 2 above. 

Comment:

5. In a mysterious departure from its normal procedures, the management
of OPP withdrew from its staff scientists the determination of an
appropriate FQPA safety factor for DDVP. Staff scientists conducted the
FQPA safety factor analysis and made decisions for all but one of the 40
organophosphates: DDVP. Inexplicably, OPP placed the FQPA safety factor
analysis and decision for DDVP before the OPP Division Directors. We are
unaware of any other instance in the public record of RED risk
assessments in which a scientific assessment and decision of this nature
was placed in the hands of a group of managers, most of whom who have
neither the training nor qualifications to make the toxicological and
exposure assessments needed for this decision. As EPA explains: “The
toxicology database was evaluated for the neurotoxic, developmental and
reproductive toxic potential of the 40 organophosphates . . . . For one
chemical (Dichlorvos or DDVP), the FQPA Safety Factor was determined by
the Division Directors.” (HED Risk Assessment at 27). No explanation
was given for this extraordinary action. The resulting decision to
remove most of the presumptive tenfold safety factor was a flawed and
unsupportable decision.

EFED Response:

This comment does not relate to the ecological risk assessment.  It will
be addressed in the IRED response to comments.  

Comment:

6. Any meeting that includes discussion of issues pertinent to an
ongoing Special Review is governed by the Special Review rules set forth
at 40 C.F.R. § 154.27(b), which, among other things, prohibit EPA from
committing to take any particular action in such meetings. Yet in these
meetings, EPA agreed to a laundry list of concessions to Amvac. Other
secret meetings between EPA and Amvac have never been memorialized as
required by the Special Review regulations. (40 C.F.R. § 154.15(b)(7)).
Each such meeting should have been memorialized and docketed within 10
working days of the meeting. (Id. § 154.15(d)). This provision has been
repeatedly violated by the Agency. Even when the Agency has docketed
memoranda summarizing meetings with Amvac, these memoranda contain
little useful information and fail to comply with the detailed content
requirements for docketing such meetings (Id. § 154.15(b)(7)). The
Agency must immediately supplement the existing memoranda and create and
docket complying memoranda for all of the undocketed meetings with Amvac
concerning DDVP.

EFED Response:

This comment does not relate to the ecological risk assessment.  It will
be addressed in the IRED response to comments.  

Comment:

7. As detailed in the specific comments below, there are no DDVP uses
for which the database is adequate to meet the legal standards for
continued registration or tolerances.  In addition, the available data
demonstrate that all uses of DDVP are in fact unsafe and are likely to
cause unreasonable adverse effects on the environment. In many
instances, EPA even acknowledges these unreasonable risks. Accordingly,
all uses of DDVP should be cancelled and all tolerances revoked.

EFED Response:

This comment will be addressed in the responses to the specific comments
below.

Comment:

8. There is nothing in the record of this RED to indicate that EPA has
sought to empirically validate the efficacy and successful use of any of
the many risk mitigation measures that it might impose as part of a
final decision on DDVP. Thus, measures such as buffer strips,
requirements for wearing protective clothing and equipment, and detailed
use instructions need to be shown not only to be capable of reducing
risk to the degree assumed by the Agency, but also to be understood,
accepted, and successfully used by users. Even the most efficacious risk
reduction measure is useless if it will not be understood or followed by
users. The Agency must undertake a large scale study of the measures it
commonly imposes and relies upon to determine if they are efficacious
and if they are generally followed successfully by users. In the absence
of empirical proof of the efficacy and widespread and generally accepted
successful use of these measures, they cannot be relied upon to maintain
any registrations or tolerances in effect.

EFED Response:

This comment does not relate to the ecological risk assessment.  It will
be addressed in the IRED response to comments.  

Comment:

9. In at least one instance in the record, it is clear that dissenting
views were expressed by staff members in an important decision meeting,
but the substance of these views was not reported in the memorandum of
the meeting. (CARC Final Report, Feb. 2, 2000). The record of DDVP and
all other chemicals subject to regulation must be supplemented with
detailed information concerning the substance of all dissenting views
expressed by EPA staff. Failure to fully and frankly describe dissent is
arbitrary and capricious and undermines the credibility of the Agency
and its decisions.

EFED Response:

This comment does not relate to the ecological risk assessment.  It will
be addressed in the IRED response to comments.  

Comment:

10. The Agency identifies many instances in which labels are
inconsistent with registration requirements or do not contain limits on
use, proper protective measures, or warnings. The Agency must explain to
the public how these labels could remain on the market during the
previous 17 year scrutiny of DDVP. In any case, all DDVP labels must be
thoroughly reviewed and brought into compliance with statutory and
regulatory requirements and made consistent with the requirements of
their registrations.

EFED Response:

This comment does not relate to the ecological risk assessment.  It will
be addressed in the IRED response to comments.  

Comment:

11. The Agency identifies several DDVP degradates: 2,2-dichloroacetic
acid (DCA), 2,2-dichloroacetaldehyde, 2,2-dichloroethanol (DCE), and
desmethyl dichlorvos (methyl O-(2,2-dichlorovinyl) phosphate). It is not
clear that these degradates were ever searched for or even detectable in
the various monitoring and metabolism studies relied upon by the Agency.
For every study relied upon by the Agency in the DDVP record in which
analytical chemistry was performed, the Agency should supplement the
record with information describing whether the analytical method used
was capable of identifying the degradates of DDVP, whether the
investigators looked for the degradates, and the levels of detection and
quantitation for each of the degradates in each study. There is no
indication whether these degradates were ever separately subjected to
toxicological testing. Thus, the Agency, in addition to not having
environmental and dietary monitoring data for the degradates, may also
be lacking basic human and wildlife toxicology data concerning these
degradates. These data should have been required to be submitted during
the 17 year Special Review process and the RED process. In light of the
examples of pesticide degradates that are more toxic and persistent than
the parent compounds (e.g., for the organochlorine pesticides), the
incomplete database for these degradates makes it impossible for EPA to
find that DDVP is safe and will not cause unreasonable adverse effects.
Accordingly, all registrations for DDVP must be cancelled and all
tolerances revoked.

EFED Response:

The above-mentioned degradates were indeed looked for and quantified in
the laboratory environmental fate studies as well as in the terrestrial
field dissipation studies (see pages 20-22 of the ecological risk
assessment).  A good material balance was reported in the field study. 
As indicated in the residue chemistry chapter, it was determined that
the degradates are much less toxic than the parent compound, therefore,
the only residue for toxicological concern for mammals was DDVP.   There
is also no evidence pointing to a concern about the degradates for other
taxa. 

Comment:

12. The basic toxicology data for DDVP does not include full testing of
end use products or separate testing of inert ingredients and impurities
in the end-use products. The monitoring studies for residues of DDVP do
not appear to have included testing for residues of inert ingredients
and impurities. There is no indication in the DDVP record that EPA has
complete data on the toxicology and residue distribution of all of the
inerts and impurities in all of the DDVP end-use products. As EPA is
well aware, many substances that are legally classified in FIFRA as
“inerts” are not, in fact, biologically or chemically inert. Indeed,
in one of the few instances in which DDVP data were available for both
an end use and a technical product, the end use product was found to be
tenfold more toxic than the technical product. (Revised Ecological Risk
Assessment at 29). Since the only difference between the end use and
technical products is the inert ingredients in the end use product, the
results of these studies suggest that the inerts in the tested product
were far more toxic than the active ingredient. Accordingly, the Agency
needs to provide the public with detailed information about its review
of the risks posed by the inert ingredients and impurities in all DDVP
end-use products.

The Confidential Business Information (CBI) provisions of FIFRA are not
an excuse for not providing this information. The Special Review and RED
documents are “findings of fact” within the meaning of FIFRA Section
10(b), and thus may include disclosure of formulation information
otherwise protected from disclosure under Section 10. In any case, there
is no basis for keeping secret the extent of the efforts made by the
Agency to assess the risks of the inerts and impurities in DDVP end-use
products. If the Agency has not fully assessed the risks of these
components of DDVP end-use products, then it is impossible for the
Agency to conclude that these products are safe and will not cause
unreasonable adverse effects on the environment. Accordingly, all
registrations for DDVP would have to be cancelled and all tolerances
revoked.

EFED Response:

Risk assessments performed for purposes of reregistration focus on the
risk of the active ingredient under the assumption that the active
ingredient is the main causative agent of toxicity.  Long experience has
provided a basis for concluding that toxicity data on the active
ingredient is sufficient to evaluate the risks of products containing
that active ingredient.  Therefore, except as noted below, EPA does not
routinely require testing of either the inerts or the formulations.

The Agency's product reregistration phase captures information about the
acute risks attributable to the formulation and sets appropriate
restrictions (labeling usually) that reflects those risks.  The Agency
also has a program that looked at food use inerts under FQPA.  Finally,
there is an ongoing program that involves looking at other inerts in a
tiered approach that focuses on those inerts which are known to be toxic
or which are structurally related to known toxicants.

To the extent that EPA has data showing that risks from exposure to the
end-use formulation of an active ingredient are higher than would be
expected based on the amount of the active ingredient in the
formulation, that data is reported in the risk assessment.  In the
instant case, the test data that indicated a ten-fold increase in
toxicity for one of the end use products did not affect the conclusions
of the risk assessment of the active ingredient because the species
tested (oyster) was determined to not be the most sensitive species, and
thus that test data was not used to set the acute aquatic invertebrate
endpoint for DDVP.  Nevertheless, that test data will be a prime
consideration when the Agency determines what further product-specific
testing will be required in  the data call-in.     

Comment:

13. In conjunction with the 1995 PD-2/3, 60 Fed. Reg. 50337, EPA
identified a number of registrations that would be cancelled or amended.
The Agency also issued a notice announcing its intention to cancel or
amend these registrations. (60 Fed. Reg. 19580).

However, a random check of the labels for these products on EPA’s
Pesticide Product Labeling System (PPLS) revealed that, for all of the
sampled labels, none of these cancellations or amendments had been
posted on PPLS. Either PPLS is very outdated or EPA has failed to
implement its decision to cancel or amend these products. In either
case, EPA must explain the status of these actions and explain whether
the PPLS reports are accurate. If the cancellations have not been
implemented by means of issuance of final cancellation orders or if the
proposed label changes have not been made, then EPA must explain its
reasons for not implementing these decisions, and all of the risk
assessments must be amended to include risks from uses that should have
been cancelled, deleted, or amended but still remain unchanged or on the
market.

EFED Response:

This comment does not relate to the ecological risk assessment.  It will
be addressed in the IRED response to comments.  

SPECIFIC COMMENTS

Revised Ecological Risk Assessment (RERA), June 20, 2005:

Comment:

Pages 8-9: The RERA is not consistently based on maximum label rates.
Indeed, it appears some labels do not specify limits on use. RERA at 9
(“many of the labels have vague directions for use.”). Default
assumptions with such labels are not practical because any estimate may
be an underestimate. All DDVP labels that do not specify the maximum
frequency of treatments, the areas where treatments can be performed,
minimum precautions and protective clothing and equipment, maximum
concentrations, and appropriate application equipment and parameters
should be amended as necessary or cancelled.

EFED Response:

While the risk assessment assumed 75 applications per year for the
flying insect scenario and 120 applications per year for the granular
bait, the estimated exposure values would change very little, if at all,
if there were even more applications because the degradation and
dissipation of DDVP are so rapid.  DDVP has a foliar dissipation half
life of only 2 hours and a soil metabolism half life of 10 hours.  Thus,
there is little or no accumulation from application to application.  The
RQ’s would be essentially the same no matter how many applications
were assumed.  The only difference would be the probability of contact
with a non-target organism.  The more often the chemical is applied, the
greater the chance for exposure.  But, the overall conclusions of the
screening level risk assessment remain the same.

Comment:

Page 9: Basic, reliable information concerning use of DDVP, such as
whether DDVP is used for mosquito abatement, is apparently not always
available: “Dichlorvos does not appear to be used in this country for
adult mosquito control. It is not listed in State Management
recommendations for mosquito control, and the American Mosquito Control
Association (AMCA) has indicated as far as they could tell, it wasn’t
being used in this country.” There is no way for the Agency to protect
the environment from potentially dangerous exposures to DDVP if it does
not have reliable information about where and how the product is used.
In the case of the mosquito control use, if the principal user groups
neither recommend nor are aware of its use, it should be cancelled. If
this use were continued it should be required to be packaged as a single
use product to allow direct monitoring of the total amount used each
year via FIFRA Section 7 reports.

EFED Response:

Whether or not DDVP is actually being used for mosquito abatement, the
risk assessment includes a flying insect scenario which covers the
potential mosquito use.

Comment:

Page 9: Use information about turf applications is based on what appears
to be anecdotal, unverified information: “For turf use, dichlorvos
would normally be applied only once or twice per season. It is possible
that up to four applications may be made, but this would be unusual
(Douglas Sutherland, BEAD entomologist, personal communication,
4/13/98).” In an effort to identify a “high end” estimate, the
Agency offers the following: “However, since the label does not limit
the number of applications, the high end estimate of 4 applications per
season is modeled in addition to 1 application per season.” There is
no indication that this estimate is based on anything other than pure
guesswork. Guesses are not an acceptable basis for regulatory
decisionmaking. The Agency should have required the registrants to
develop reliable use information in the RED process. Any turf labels
that do not limit the frequency of applications should be immediately
amended to impose such restrictions or the products should be cancelled.

EFED Response:

Turf use of DDVP is being canceled.

Comment:

Page 9: Without explanation, the Agency failed to conduct a literature
survey: “for this risk assessment for dichlorvos, a detailed open
literature search was not conducted.” It is difficult to imagine a
justification for this deliberate failure to search for readily
available, relevant scientific information. The failure to do so is
arbitrary and capricious, and a detailed literature survey must be
conducted before DDVP may be reregistered.

EFED Response: 

The registrant has submitted studies to satisfy all data requirements
listed in 40 CFR Part 158.  No major data gaps were identified in the
risk assessment.  Below is a bibliography of the registrant-submitted
studies used in the risk assessment:

Grimes, J.; Jaber, M. (1988) DDVP: An Acute Oral Toxicity Study with the
Bobwhite: Final Report: Project No. 246-102. Unpublished study prepared
by Wildlife International Ltd. 21 p.

Hudson, R.; Tucker, R.; Haegele, M. (1984) Handbook of toxicity of
pesticides to wildlife: Second edition. US Fish and Wildlife Service:
Resource Publication 153. 91 p.

Hill, E.F.; Heath, R.G.; Spann, J.W.; et al. (1975) Lethal Dietary
Toxicities of Environmental Pollutants to Birds: Special Scientific
Report--Wildlife No. 191. (U.S. Dept. of the Interior, Fish and Wildlife
Service, Patuxent Wildlife Research Center; unpublished report)

Cameron, D. (1996) DDVP: Bobwhite Quail Dietary Reproduction and
Tolerance Studies: Lab Project Number: AVC 2: AVC 2/960139. Unpublished
study prepared by Huntingdon Life Sciences Ltd. 324 p.

Redgrave, V. (1997) Mallard Duck Dietary Reproduction and Tolerance
Studies: DDVP: Lab Project Number: AVC 7/961821: AVC 7. Unpublished
study prepared by Huntingdon Life Sciences Ltd. 243 p.

Gaines, T.B. (1960) The acute toxicity of pesticides to rats. Toxicology
and Applied Pharmacology 2(1):88-99. (Also in unpublished submission
received Feb 14, 1961 under PP0299; submitted by Chemagro Corp., Kansas
City, Mo.; CDL:090320-Q)

MacDonald, R.; Thorpe, E.; Hendy, R.; et al. (1982) Toxicology of
Consumer Products: The Acute (4 H) Inhalation Toxicology of Dichlorvos
Vapour in Rats and Mice: External Report SBER.82.008. (Unpublished study
received Jan 24, 1984 under 201-125; prepared by Shell Research Ltd.,
Eng., submitted by Shell Chemical Co., Washington, DC; CDL:252309-A)

Tyl, R.; Myers, C.; Marr, M. (1992) Two-Generation Reproductive Toxicity
Study of DDVP Administered in Drinking Water to CD (Sprague-Dawley)
Rats: Final Report: Lab Project Number 60C-4629-170. Unpublished study
prepared by Research Triangle Institute. 1225 p.

Atkins, E.L.; Greywood, E.A.; Macdonald, R.L. (1975) Toxicity of
Pesticides and Other Agricultural Chemicals to Honey Bees: Labo- ratory
Studies. By University of California, Dept. of Entomology. ?: UC,
Cooperative Extension. (Leaflet 2287; published study.)

Mayer, F.; Ellersieck, M. (1986) Manual of Acute Toxicity:
Interpretation and Data Base for 410 Chemicals and 66 Species of
Freshwater Animals. US Fish & Wildlife Service, Resource Publication
160. 579 p.

Jones, F. (1994) DDVP 4-E Emulsifiable Concentrate: Acute Toxicity to
Rainbow Trout (Oncorhynchus mykiss) (sic) under Flow-Through Test
Conditions: Lab Project Number: J9403007E. Unpublished study prepared by
Toxikon Environmental Sciences. 49 p.

Johnson, W.; Finley, M. (1980) Handbook of Acute Toxicity of Chemicals
to Fish and Aquatic Invertebrates: Resource Publication 137. US Fish and
Wildlife Service, Washington, D.C. 106 p.

Jones, F. (1994) DDVP 4-E Emulsifiable Concentrate: Acute Toxicity to
Bluegill (Lepomis machrochirus) under Flow-Through Test Conditions: Lab
Project Number: J9403007D. Unpublished study prepared by Toxikon
Environmental Sciences. 49 p.

Davis, J. (1995) DDVP Technical Grade: Toxicity to Embryos and Larvae of
the Rainbow Trout, Oncorhynchus mykiss, Under Flow-Through Conditions:
Lab Project Number: J9403007M. Unpublished study prepared by Toxikon
Environmental Sciences. 77 p.

Ward, G.; Davis, J. (1995) DDVP Technical Grade: Chronic Life-Cycle
Toxicity to the Water Flea, Daphnia magna, Under Flow-Through Test
Conditions: Lab Project Number: J94030071: J9403007N. Unpublished study
prepared by Toxikon Environmental Sciences. 74 p

Jones, F.; Davis, J. (1994) DDVP Technical Grade: Acute Toxicity to
Sheepshead Minnow (Cyprinodon variegatus) Under Flow-through Test
Conditions: Lab Project Numbers: J9403007F: J9403007B. Unpublished study
prepared by Toxikon Environmental Sciences. 59 p.

Jones, F.; Davis, J. (1994) DDVP 4-E Emulsifiable Concentrate: Acute
Toxicity to Sheepshead Minnow, Cyprinodon variegatus, Under Flow-through
Test Conditions: Lab Project Numbers: J9403003I: J9403007B. Unpublished
study prepared by Toxikon Environmental Sciences. 60 p.

Ward, G.; Davis, J. (1995) DDVP Technical Grade: Toxicity to Embryos and
Larvae of the Sheepshead Minnow (Cyprinodon variegatus) Under
Flow-Through Test Conditions: Lab Project Number: J9407006B. Unpublished
study prepared by Toxikon Environmental Sciences. 81 p.

Jones, F.; Davis, J. (1995) DDVP Technical Grade: Acute Effect on New
Shell Growth of the Eastern Oyster (Crassostrea virginica): Lab Project
Numbers: J9403007H: J9403007B. Unpublished study prepared by Toxikon
Environmental Sciences. 63 p.

Jones, F.; Davis, J. (1994) DDVP 4-E Emulsifiable Concentrate: Acute
Effect on New Shell Growth of the Eastern Oyster (Crassostrea
virginica): Lab Project Numbers: J9403007K: J9403007C. Unpublished study
prepared by Toxikon Environmental Sciences. 58 p.

Jones, F.; Davis, J. (1994) DDVP Technical Grade: Acute Toxicity to the
Mysid (Mysidopsis bahia) Under Flow-through Test Conditions: Lab Project
Numbers: J9403007G: J9403007B. Unpublished study prepared by Toxikon
Environmental Sciences. 61 p.

Jones, F.; Davis, J. (1994) DDVP 4-E Emulsifiable Concentrate: Acute
Toxicity to the Mysid, Mysidopsis bahia, Under Flow-through Test
Conditions: Lab Project Numbers: J9403007J: J9403007B. Unpublished study
prepared by Toxikon Environmental Sciences. 60 p.

Ward, S.; Davis, J. (1995) DDVP Technical Grade: Chronic Toxicity to the
Mysid (Mysidopsis bahia) Under Flow Through Conditions: Lab Project
Number: J9407006A. Unpublished study prepared by Toxikon Environmental
Sciences. 90 p.

Vithala, R. (1990) DDVP--Hydrolysis of C14| DDVP in Aqueous Solutions
Buffered at pH 5, 7, and 9: Lab Project Number: 006/006/ 001/89.
Unpublished study prepared by Univ. of Pittsburgh, Center for Hazardous
Materials Research. 141 p.

Mobley, S. (1994) Aqueous Photolysis of (carbon 14) DDVP (Dichlorvos) in
Artificial Light: Lab Project Number: 838: 1611. Unpublished study
prepared by PTRL East, Inc. 95 p.

Misra, B. (1995) Photodegradation of (Carbon-14)-DDVP (Dichlorvos) on
Sandy Loam Soil Under Artificial Sunlight Irradiation: Final Report: Lab
Project Number: ME 9400184. Unpublished study prepared by Pittsburgh
Environmental Research Lab, Inc. 79 p

Vithala, R. (1990) DDVP--Aerobic Soil Metabolism of C 14| DDVP: Lab
Project Number: 006/006/004/89. Unpublished study prepared by Univ. of
Pittsburgh, Center for Hazardous Materials Research. 7 p.

Mobley, S. (1995) Anaerobic Soil Metabolism of (carbon 14)DDVP
(Dichlorvos): Lab Project Number: 882: 1847. Unpublished study prepared
by PTRL East, Inc. 82 p.

Vithala, R. (1990) DDVP--Mobility of Unaged DDVP Using Batch Equilibrium
Technique: Lab Project Number: 006/006/008/89. Unpublished study
prepared by Univ. of Pittsburgh, Center for Hazardous Materials
Research. 70 p

Leppert, B. (1997) Dichlorvos: Terrestrial Field Dissipation of DDVP 2-E
on Bareground Plot in Missouri: Final Report: Lab Project Number:
SARS-96-MO-71: CHW 6293-147: SARS-96-71. Unpublished study prepared by
Stewart Agricultural Research Services, Inc. and Corning Hazleton, Inc.
440 p.

Leppert, B. (1997) Dichlorvos: Terrestrial Dissipation of DDVP 2-E on
Bareground Plot in California...: Final Report: Lab Project Number:
6293-150: SARS-96-CA-72: SARS-96-72. Unpublished study prepared by
Stewart Agricultural Research Services, Inc. and Covance Lab., Inc. 404
p

Leppert, B. (1997) Dichlorvos: Terrestrial Field Dissipation of DDVP 2-E
on Bareground Plot in Missouri: Revised Final Report: Lab Project
Number: SARS-96-MO-71: CHW 6293-147: SARS-96-71. Unpublished study
prepared by Stewart Agricultural Research, Inc. and Corning Hazleton,
Inc. 445 p.

Pack, D. (1986) Freundlich Adsorption Isotherms of Dichloracetic Acid:
Laboratory Project Identificaton: 8614820. Unpublished study prepared by
Chevron Chemical Co. 15 p.

Since late 2004, OPP has been using the ECOTOX database on a routine
basis as a search engine to identify open literature studies that may
potentially be used in ecological risk assessments.  However, at the
time the DDVP risk assessment was done, OPP was not routinely performing
ECOTOX-based literature searches, especially when no major data gaps had
been identified in the required Part 158 information submissions.

An ECOTOX search will be scheduled for DDVP during Registration Review
and the risk assessment will be updated as necessary.

Comment:

Pages 10-11: The RERA indicates that labels lack necessary limitations
on applications: “Currently registered labels for turf and flying
insects allow for fogging and misting, and there are no label
prohibitions against aerial application. Labels do not specify maximum
numbers of applications or reapplication intervals.” Despite the
absence of limitations, the Agency makes a number of “worst case
assumptions” that appear to be based on anecdotal, unverified
information. Clearly, it is not possible to determine what constitutes a
worst case scenario when there are no restrictions on use. EPA’s
speculative approach in this regard is unjustified.

EFED Response:

While the risk assessment assumed 75 applications per year for the
flying insect scenario, the estimated exposure values would change very
little, if at all, if there were even more applications because the
degradation and dissipation of DDVP are so rapid.  DDVP has a foliar
dissipation half life of only 2 hours and a soil metabolism half life of
10 hours.  Thus, there is little or no accumulation from application to
application.  The RQ’s would be essentially the same no matter how
many applications were assumed.  The only difference would be the
probability of contact with a non-target organism.  The more often the
chemical is applied, the greater the chance for exposure.  But, the
overall conclusions of the screening level risk assessment remain the
same.

Comment:

Page 16: Although the Agency concludes that it is missing needed
terrestrial plant data, it did not impose a data requirement to obtain
this information: “There are no terrestrial plant data for dichlorvos
which leads to uncertainty in the evaluation of plant risk and indirect
effects to other organisms.” The Agency gives no reason for its
decision not to require this data. This data should have been required
to be submitted during the RED process. In the absence of adequate
terrestrial plant data, all terrestrial uses of DDVP should be canceled.

EFED Response:

Plant testing is generally only required for herbicides and fungicides. 
For other pesticides, such as insecticides, e.g., DDVP, the plant
testing data requirement is triggered when there is sufficient
information or reason to believe the chemical may be phytotoxic.  For
DDVP, there is no evidence leading us to believe that this compound
could have adverse effects on plants.  Consistent with 40 CFR Part 158,
plant testing is not required for DDVP.

Comment:

Page 18: The Agency states that the “rationale for choosing four
applications for turf was based on information received from BEAD
indicating a worst-case scenario would probably be about four
applications.” There is no indication that this BEAD information is
anything other than unverified, anecdotal information. The use of such
information is subject to the comments and criticisms set forth for
pages 8-9, above.

EFED Response:

The use of DDVP on turf is being canceled.

Comment:

Pages 21-22: The RERA identifies several DDVP environmental degradates:
2,2- dichloroacetic acid (DCA), 2,2-dichloroacetaldehyde (DAA),
2,2-dichloroethanol (DCE), and desmethyl dichlorvos (methyl
O-(2,2-dichlorovinyl) phosphate). These degradates do not appear to be
detectable by the methods used in the available water monitoring
studies.  It is not clear whether the analytical methodology used in
other environmental monitoring studies were capable of detecting these
degradates. Indeed, it is not clear whether food and drinking water
exposure studies were capable of detecting these degradates. The RED
process should have required proper monitoring capable of detecting all
of DDVP’s degradates below toxicologically significant levels. In the
absence of such data, all registrations of DDVP should be cancelled and
all tolerances revoked.

EFED Response:

As stated above, the degradates were indeed looked for and quantified in
the laboratory environmental fate studies as well as in the terrestrial
field dissipation studies (see pages 20-22 of the ecological risk
assessment).  A good material balance was reported in the field study. 
As indicated in the residue chemistry chapter, it was determined that
the degradates are much less toxic than the parent compound, therefore,
the only residue for toxicological concern for mammals was DDVP.   There
is also no evidence pointing to a concern about the degradates for other
taxa. 

Comment:

Page 22: The RERA uses models to predict exposure: “Tier II Estimated
Environmental Concentrations (EECs) for dichlorvos for the turf
scenarios were estimated using EFED’s aquatic models PRZM-EXAMS
(EXposure Analysis Modeling System). PRZM is used to simulate pesticide
transport as a result of runoff and erosion from an 10-ha agricultural
field, and EXAMS considers environmental fate and transport of
pesticides in surface water and predicts EECs in a standard pond.”
There is no indication in the DDVP record that any of these models has
been validated by means of empirical data from studies designed to test
the reliability of these models. In the absence of validated models and
reliable empirical data, the Agency cannot conclude that registrants
have met their burden of proving that DDVP will not cause unreasonable
adverse effects on the environment. In addition, rather than using
speculative assumptions and unvalidated models, the Agency should have
required the registrants to conduct and submit appropriate studies
sufficient to resolve all of the uncertainties now addressed by these
models. In light of the inadequacy of the database for environmental
exposures, all registrations for outdoor uses of DDVP should be
cancelled.

EFED Response:

As explained above, the PRZM EXAMS model has been extensively peer
reviewed and validated. 

Comment:

Page 24: The Agency resorts to speculation in its assessment of risks to
aquatic life: “For the flying insect (including adult mosquitoes) use,
EFED currently has no models that would be appropriate for modeling
EECs. PRZM/EXAMS and the GENEEC model are inappropriate to use. It is
likely EECs found in surface water from treatment for flying insects
(including adult mosquitoes) would likely be lower than EECs from
treatment to turf, since the treatment area would likely be less. Since
the applications for flying insect control are ground applications
(e.g., back-pack sprayers or truck-mounted sprayers) using coarse sprays
directed to the vegetation (no fogging or misting), EFED cannot perform
a quantitative assessment.” Although ultimately the Agency decides
that it does not have an adequate basis for a quantitative assessment,
it does not conclude, as FIFRA requires, that the uses at issue have not
been shown not to cause unreasonable adverse effects on the environment.
Accordingly, these uses should be cancelled.

EFED Response:

Concentrations expected in the aquatic environment from the flying
insect control use are predicted to be lower than from the use on turf,
due to the much larger treatment areas for the turf scenario. 
Therefore, aquatic risk from the flying insect control use will be lower
than what was predicted for the turf use (see pages 37-39 of the
ecological risk assessment

The Agency is proceeding with its assessment of potential risks from the
use of DDVP to federally listed threatened and endangered species,
during which time the exposure analysis will be refined.

Comment:

Page 24: The RERA uses another model for liquid and granular
applications: “For obtaining EECs for acute exposure from multiple
applications and chronic exposure from both single and multiple
applications of liquid dichlorvos and granular bait products, the TREX v
1.1 (U.S. EPA. 2004b) program was used.” The cited document does not
appear to be in the DDVP record, therefore it is not possible to
determine whether the model has been validated with empirical evidence
from a study designed to test the reliability of the model. There is
also no indication why monitoring data was not required to be submitted
by registrants. In the absence of evidence that the T-Rex model has been
validated and is reliable, it is not possible for the Agency to conclude
that the uses in question will not cause unreasonable adverse effects on
the environment. Accordingly, these uses should be cancelled.

EFED Response:

For spray applications, estimation of pesticide concentration in
wildlife food items focuses on quantifying possible dietary ingestion of
residues on vegetative matter and insects.  The residue estimates are
based on a nomogram that relates food item residues to pesticide
application rate.  The nomogram is based on an EPA database called UTAB
(Uptake, Translocation, Accumulation, and Biotransformation), a
compilation of actual measured pesticide residue values on plants
(Nellessen, J.E.,and J.S. Fletcher. 1992. UTAB: A computer database on
residues of xenobiotic organic chemicals and heavy metals in plants. J.
Chem. Inf. Comput. Sci. 32:144-148.), and work from Fletcher et al.
(Fletcher, J.S., J.E. Nellessen, and T.G. Pfleeger. 1994. Literature
review and evaluation of the EPA food-chain (Kenaga) nomogram, an
instrument for estimating pesticide residues on plants. Environmental
Toxicology and Chemistry 13(9):1383-1391.)

For granular, bait, and treated seed applications, estimation of
loadings of pesticide per unit area are calculated.  This approach,
which is intended to represent exposure via multiple routes and not just
direct ingestion, considers observed effects in field studies and
relates them to the pesticide applied to surface area.

T-REX is not a model per se, but rather it is a spreadsheet tool that
automates the risk quotient calculations using the residue estimates
described above and the toxicity data.  It does not represent a new or
different terrestrial exposure method.  More information about the
calculations performed in T-REX may be found at:

  HYPERLINK
"http://www.epa.gov/oppefed1/ecorisk_ders/toera_analysis_exp.htm#HDEEEOP
" 
http://www.epa.gov/oppefed1/ecorisk_ders/toera_analysis_exp.htm#HDEEEOP 

The T-REX user’s guide may be found at:

  HYPERLINK "http://www.epa.gov/oppefed1/models/terrestrial/index.htm" 
http://www.epa.gov/oppefed1/models/terrestrial/index.htm 

Comment:

Page 24: The Agency uses less than maximum label rates based on
unsubstantiated anecdotal information for turf use: “For the liquid
spray application to turf, the maximum application rate modeled was 0.2
lb ai/A. One application and four applications (with 30 day application
interval) were modeled for turf. The rationale for choosing four
applications for turf was based on information received from BEAD
indicating a worst case scenario of four applications.” Similarly,
according to EPA, “The rationale for choosing weekly applications for
mosquito control was based on information received from BEAD indicating
a worst case scenario for adult mosquito control would probably be
around 75 applications to a given site over a year period.” There is
no indication where BEAD obtained its information in these two cases. If
the information is from unverified anecdotal sources, it should be
rejected. Even if it could be empirically demonstrated that most uses
are consistent with BEAD’s estimates, the fact remains that a
significant subgroup of users might apply the product more frequently.
In the absence of label limitations on the maximum number of
applications, it is impossible to even establish a worst case scenario.
If, as is the case here, the risk cannot be bounded by a default
assumption, the relevant labels must be amended to impose a safe limit
on frequency of applications, or the use must be cancelled.

EFED Response:

The registrant has requested voluntary cancellation of the use on turf;
therefore our response will be limited to the flying insect scenario. 
While the risk assessment assumed 75 applications per year, the
estimated exposure values would change very little, if at all, if there
were even more applications because the degradation and dissipation of
DDVP are so rapid.  DDVP has a foliar dissipation half life of only 2
hours and a soil metabolism half life of 10 hours.  Thus, there is
little or no accumulation from application to application.  The RQ’s
would be essentially the same no matter how many applications were
assumed.  The only difference would be the probability of contact with a
non-target organism.  The more often the chemical is applied, the
greater the chance for exposure.  But, the overall conclusions of the
screening level risk assessment remain the same.

Comment:

Page 26: Although the Agency appears to recognize that the wildlife
toxicity data it has is not representative, it nonetheless performs a
broad extrapolation anyway: “Toxicity testing reported in this section
does not represent all species of birds, mammals, or aquatic organisms.
Only a few surrogate species for both freshwater fish and birds are used
to represent all freshwater fish (2000+) and bird (680+) species in the
United States.  Mammalian acute studies are usually limited to Norway or
New Zealand rat or the house mouse. Estuarine/marine testing is usually
limited to a crustacean, a mollusk, and a fish. Also, neither reptiles
nor amphibians are tested. The risk assessment assumes that avian and
reptilian toxicities are similar. The same assumption is used for fish
and amphibians.”  The huge leap of faith implicit in extrapolating
from a handful of test species to thousands of disparate species is not
explained or supported by reference to any validating data.  Translating
data from even closely related species is a questionable practice at
best, because of the many variables; translating data across an entire
family without regard to the vast differences between genera is
scientifically unsupportable. EPA must either make a persuasive case for
the reliability of this extrapolation or it must cancel all uses of DDVP
that might result in exposure to wildlife.

EFED Response:

Rarely are toxicity data available for the specific species identified
in the risk assessment endpoints.  40 CFR Part 158 specifies the types
and amounts of data that the Agency needs to determine the risks of a
pesticide to wildlife, aquatic organisms, and plants.  In the majority
of cases, the screening-level risk assessment process relies on a suite
of toxicity studies performed on a limited number of organisms in the
following broad groupings:

Birds (mallard duck and bobwhite quail) used as surrogate for
terrestrial-phase amphibians and reptiles,

Mammals (laboratory rat),

Freshwater fish (bluegill sunfish, rainbow trout, and fathead minnow)
used as a surrogate for aquatic phase amphibians,

Freshwater invertebrates (Daphnia magna),

Estuarine/marine fish (sheepshead minnow),

Estuarine/marine invertebrates (Crassostrea virginica and Mysidopsis
bahia),

Terrestrial plants (corn, soybean, carrot, oat, tomato, onion, cabbage,
lettuce), and

Algae and aquatic plants (Lemna gibba, Skeletonema costatum, Anabaena
flos-aquae, Selenastrum capricornutum, Clorell vulgaris, Scenedesmus
subspicatus).

Within each of these very broad taxonomic groups, an acute and a chronic
endpoint are selected from the available test data.  The selection is
made from the most sensitive species tested within the organism group. 
If additional toxicity data for more species of organisms in a
particular group are available, the selection need not be limited to the
species listed above, but may be expanded to include data for other
species/studies that meet the data quality requirements.  Regardless of
the extent of data beyond the regulation-required set of toxicity
studies, the risk assessment relies on selection of endpoints from the
most sensitive species tested in acceptable studies.

This general approach to ecological risk assessment has been presented
to the FIFRA Scientific Advisory Panel and was supported by the Panel
members.

  HYPERLINK
"http://www.epa.gov/scipoly/sap/meetings/1998/december/risk.pdf" 
http://www.epa.gov/scipoly/sap/meetings/1998/december/risk.pdf 

Comment:

Page 26: EPA states that “There are no core studies available for the
rainbow trout.” Core studies are basic studies required by 40 C.F.R.
Part 158. These studies should have been required to be submitted. Any
registrations granted on the condition of submission of these studies
should be cancelled pursuant to FIFRA Section 6(e). In the absence of
core trout studies, all registrations that might result in exposure to
trout and any other species for which the trout data are used as a
surrogate should be cancelled.

EFED Response:

The Agency reviewed two rainbow trout studies.  Both studies were found
to be scientifically valid, but because of minor deviations from Agency
guidelines, the studies were classified as supplemental and not core. 
Supplemental studies are used in risk assessments.  Note that the most
sensitive endpoint for freshwater fish, an LC50 of 183 ppb for the lake
trout, was used in the DDVP assessment and this study was classified as
supplemental.

Comment:

Page 29: Oyster data suggested a disturbing phenomenon, but the Agency
failed to pursue the issue. EPA states that, “in the case of the
oyster, a large discrepancy exists, with toxicity of the EC formulation
(on an active ingredient basis) almost 10-fold greater than that of the
TGAI. No explanation for this was provided by the performing laboratory
or registrant. Since both the TGAI and TEP studies were scientifically
sound, they do not have to be repeated.” As discussed above, this
discrepancy suggests either that the inerts are very toxic or that there
was a flaw in the conduct of one of the studies. In either case, this
anomaly deserves further study. At a minimum, both studies should be
repeated and the inerts should be subjected to separate toxicology
studies.

EFED Response:

Although there is a discrepancy between the Eastern oyster technical
grade and formulated product toxicity tests, the most sensitive endpoint
(i.e., the lowest LC50 value for estuarine and marine invertebrates
tested) was for the mysid shrimp, which was used in the risk assessment
to represent potential acute risk to estuarine/marine invertebrates
(pages 29 and 38-39 of the risk assessment).  The risk quotients for
estuarine/marine invertebrates were well below Agency levels of concern.


Comment:

Page 34: The Agency used two models to estimate risk to endangered
species. There is no indication in the DDVP record that these models
have been validated by means of empirical data from studies designed to
test their reliability. Rather than using speculative assumptions and
unvalidated models, the Agency should have required the registrants to
conduct and submit appropriate studies sufficient to resolve all of the
uncertainties now addressed by these models. In the absence of
validation and reliable empirical data, the Agency cannot conclude that
registrants have met their burden of proving that DDVP will not cause
unreasonable adverse effects on the environment and that use of DDVP
will not violate the ESA. Accordingly, EPA must cancel all outdoor uses
of DDVP.

EFED Response:

The Agency risk assessment includes, as part of the risk
characterization, an evaluation of the potential probability of
individual effects for exposures that may occur at the established
endangered species level of concern.  This probability is calculated
using the established dose/response relationship and the median lethal
dose estimate for the study used to establish the toxicity endpoint for
the endangered taxa.  This approach, which is based on commonly known
and employed basic toxicology dose-response analyses, is explained on
page 64 of the Overview of the Ecological Risk Assessment Process in the
Office of Pesticide Programs (2004) document and was endorsed by the
Fish and Wildlife Service and the National Marine Fisheries Service. 
The Agency uses the Individual Effect Chance spreadsheet tool to perform
these calculations.  The information gained from this analysis is for
descriptive purposes and not for determining if there are adverse
effects.

Comment:

Page 42: The following two sentences in the RERA appear to be
contradictory:  “Honeybee acute contact toxicity values indicate that
dichlorvos is highly toxic to this insect species. Toxicity tests using
residues on foliage indicate dichlorvos is practically non-toxic to
honey bees.” EPA needs to explain this apparent contradiction.

EFED Response:

Honey bee acute contact studies represent direct contact with DDVP to
the honey bee. This situation would occur when DDVP is applied with
honey bees present in treatment areas. This is highly acutely toxic to
the honey bees, resulting in death. Foliar residue toxicity tests
involve the honey bee touching foliage after treatment. Two exposure
scenarios are considered and the most sensitive, acute contact, is used
to explain potential risk to honey bees and other beneficial insects and
pollinators.  

Comment:

Page 43: The Agency claims it could not assess risk to terrestrial
plants: “As described in the analysis section, there were no
registrant-submitted terrestrial plant studies so risk to terrestrial
plants can not be assessed.” The Agency should have required
submission of necessary studies during the RED process. Without adequate
data to assess risks to terrestrial plants, the Agency must cancel all
DDVP uses that might lead to exposure to terrestrial plants.

EFED Response:

Plant testing is generally only required for herbicides and fungicides. 
For other pesticides, such as insecticides, e.g., DDVP, the plant
testing data requirement is triggered when there is sufficient
information or reason to believe the chemical may be phytotoxic.  For
DDVP, there is no evidence leading us to believe that this compound
could have adverse effects on plants.  Consistent with 40 CFR Part 158,
plant testing is not required for DDVP.

Comment:

Page 43: Without explanation, the Agency chose not to assess mammalian
risks from baits: “Chronic risk assessments [for terrestrial mammals]
are not performed for bait products.” There is no reason to believe
that DDVP baits will pose no chronic risks to mammals. EPA’s failure
to assess this risk and failure to require submission of the necessary
data are unjustified. In the absence of a full risk assessment for
baits, the Agency cannot conclude that baits will not cause unreasonable
adverse effects on the environment, and therefore it must cancel the
bait registrations.

EFED Response:

The risk assessment includes chronic toxicity data for mammals and
chronic risk was, in fact, assessed for other uses besides the bait. 
Currently there are no approved methodologies for determining chronic
risk from a granular or bait application of a pesticide.  Therefore,
chronic risk to mammals cannot be quantified for this particular use. 
Chronic effects are described in the ecological effects characterization
on page 32. 

However, since bait uses are now limited to premise treatment around
barns, feedlots, milk sheds, poultry houses, and other farm buildings,
exposure is less likely and therefore, adverse chronic effects to birds
and mammals are not expected.

Comment:

Page 44: The Agency was also unable to assess risks to freshwater and
marine/estuarine invertebrates: “For flying insect (including adult
mosquitoes) use, EFED is unable to assess risk quantitatively. It may be
assumed that the exposure to dichlorvos from flying insect use would be
less than that expected from turf use. However, the potential risk to
freshwater and marine/estuarine invertebrates can not be quantified and
therefore can not be assessed nor discounted.” In the absence of a
full risk assessment for freshwater and marine invertebrates, the Agency
cannot conclude that DDVP registrations that might lead to exposure to
marine and aquatic invertebrates will not cause unreasonable adverse
effects on the environment, and therefore it must cancel all such
registrations.

EFED Response:

The flying insect scenario was not evaluated quantitatively because the
accepted aquatic exposure models, PRZM/EXAMS and GENEEC, are
inappropriate to use for this type of application.  Applications for
flying insect control are ground applications (e.g., backpack sprayers
or truck-mounted sprayers) using coarse sprays directed to the
vegetation.  Since the treatment area and potential for exposure would
be less for the flying insect use than for the turf use, DDVP
concentrations in surface water from treatment for flying insects would
likely be lower than the calculated EECs for turf.  We note that the
acute and chronic levels of concern for freshwater fish,
estuarine/marine fish, and estuarine/marine invertebrates were not
exceeded for the turf use.  There were however, acute and chronic LOC
exceedences for freshwater invertebrates for the turf use.

Comment:

Page 44: The Agency does not have adequate information concerning the
use of baits. It assumes small and localized use: “Exposure to aquatic
animals from bait formulations applied around animal premises is
expected to be minimal because treatment sites are small and
localized.” There is no indication that the Agency has any reliable
information to support its assumptions about bait use. On the basis of
this unsubstantiated assumption, the Agency dismisses the significance
of the risk posed by baits to aquatic species: “Therefore, the bait
formulation scenario for aquatic animals was not addressed in this risk
assessment.” It is inappropriate to ignore risk on the basis of
unsubstantiated assumptions. Furthermore, even if the Agency had a sound
empirical basis for concluding that bait use is localized and small,
even localized and small risks must be considered under FIFRA. It is
possible that small, localized exposures could cause very high localized
risks that are not justified by the concomitant benefits. In the absence
of a full risk assessment for aquatic animals, the Agency cannot
conclude that bait registrations will not cause unreasonable adverse
effects on the environment, and therefore it must cancel all bait
registrations.

EFED Response:

The bait formulation is registered for premise treatment in and around
farm buildings.  Based on the rapid soil metabolism of DDVP (page 21 of
the risk assessment) and the limited potential for this localized use to
result in contamination of water bodies, the bait formulation is not
expected to adversely impact aquatic organisms.

Comment:

Page 45: The Agency did not assess chronic risk to birds from baits:
“Chronic risk to birds from the bait formulation can not be assessed
at this time.” The necessary data to assess chronic risk should have
been required during the RED process. In the absence of reliable data
and a proper risk assessment, the Agency cannot conclude that bait
registrations will not cause unreasonable adverse effects on the
environment, and therefore it must cancel all bait registrations.

EFED Response:

The risk assessment includes chronic toxicity data for birds and chronic
risk was, in fact, assessed for other uses besides the bait.  Currently
there are no approved methodologies for determining chronic risk from a
granular or bait application of a pesticide.  Therefore, chronic risk to
birds cannot be quantified for this particular use. 

However, since bait uses are now limited to premise treatment around
barns, feedlots, milk sheds, poultry houses, and other farm buildings,
exposure is less likely and therefore, adverse chronic effects to birds
and mammals are not expected.

	

Comment:

Page 46: The discussion regarding the determination required to be made
under the Endangered Species Act implies that the Agency does not intend
to make the determinations and initiate consultation required by the
Endangered Species Act (ESA) as part of the RED. To the extent that the
Agency has made determinations in the risks assessments that endangered
species may be affected by use of DDVP products, it must immediately
begin consultation with the appropriate federal agencies. Failure to do
so violates the ESA.

EFED Response:

Please refer to the Agency's July 24, 2006 letter to Patti Goldman,
Earthjustice, for a response to this comment.  This letter has been
added to the DDVP docket.   

Comment:

Page 46: The Agency relies on its Environmental Species Protection
Program (ESPP) to justify its failure to act to protect endangered
species from the acknowledged risks of DDVP use. The ESPP provides no
legal basis for delaying compliance with the ESA.  Furthermore, the
program itself appears to have no function other than as a cover for
delaying action. It has been in place for over 16 years, yet it is still
described as “interim”: “The Endangered Species Protection Program
as described in a Federal Register notice (54 FR 27984-28008, July 3,
1989) is currently being implemented on an interim basis.” The
operative part of the program, the county bulletins, is, in every
important respect, entirely voluntary and has never been shown to be of
any value whatsoever: “As part of the interim program, the Agency has
developed County Specific Pamphlets that articulate many of the specific
measures outlined in the Biological Opinions issued to date. The
Pamphlets are available for voluntary use by pesticide applicators on
EPA’s website at www.epa.gov/espp.” The Agency has never tried to
validate the efficacy of the program as a protective measure for
endangered species. In the absence of empirical data from a study
designed to test the impact of the ESPP, it cannot be assumed to have
accomplished any degree of protection of endangered species. During the
course of the 16 years of “interim” implementation, the Agency has
had a more than adequate opportunity to determine whether the ESPP is of
any value. That it has failed to do so means that there is no basis for
concluding that the ESPP can be used to achieve compliance with the
requirements of the ESA.

EFED Response:

Please refer to the Agency's July 24, 2006 letter to Ms. Patti Goldman,
Earthjustice, for a response to this comment.  This letter has been
added to the DDVP docket.   

 

Comment:

Page 47: The Agency failed to require proper documentation of an
important study: “Raw data is not provided in the daphnid acute EC50
study (MRID 40098001/ Mayer and Ellersieck 1986) to calculate a
slope.” Because of the inadequacy of the documentation, the Agency
used a default assumption for the slope. There is no indication in the
DDVP record that there is an empirical basis for the default assumption.
There is also no indication in the record why the Agency failed to
obtain the missing raw data or why it did not require a new, properly
documented study. The same criticism applies to the bird and mammal
assessments: “Raw data are not provided in the mallard duck acute LD50
study;” “Raw data are not provided in the rat acute LD50 study (MRID
0005467) to calculate a slope.” In the absence of reliable, properly
documented data or validated default assumptions, the Agency cannot
conclude that uses that result in aquatic and terrestrial exposures will
not cause unreasonable adverse effects on the environment, and
accordingly all such registrations must be cancelled.

EFED Response:

The Agency risk assessment includes, as part of the risk
characterization, an evaluation of the potential probability of
individual effects for exposures that may occur at the established
endangered species level of concern.  This probability is calculated
using the established dose/response relationship and the median lethal
dose estimate for the study used to establish the toxicity endpoint for
the endangered taxa.  The Agency uses the Individual Effect Chance
spreadsheet tool to perform these calculations.  

Although it is preferable to have access to the raw data for analysis by
Agency scientists, this information is not always readily available. 
However, in this particular case, please recognize that the information
gained from the probability analysis is for descriptive purposes and not
for determining if there are adverse effects.

Comment:

Page 47: The probit slope analysis model used by the Agency employs what
appear to be unsubstantiated assumptions concerning dose response
slopes: “Based on an assumption of a probit dose response relationship
with a mean estimated slope of 4.5, the corresponding estimated chance
of individual mortality associated with the mammalian listed species LOC
of 0.1 is 1 in 294,000.” Extrapolations based on unsubstantiated
assumptions piled upon more unsubstantiated assumptions are inherently
unreliable. The Agency admits as much: “It is recognized that
extrapolation of very low probability events is associated with
considerable uncertainty in the resulting estimates . . . [S]ince slope
is based on a default assumption of 4.5, the 95 percent confidence
intervals for the slopes are unavailable.” Furthermore, there is no
indication in the DDVP record that this model has been validated by
means of empirical data from studies designed to test the reliability of
these models. Rather than using speculative assumptions and unvalidated
models, the Agency should have required the registrants to conduct and
submit appropriate studies sufficient to resolve all of the
uncertainties now addressed by these models. In the absence of
validation and reliable empirical data, the Agency cannot conclude that
registrants have met their burden of proving that DDVP will not cause
unreasonable adverse effects on the environment, and accordingly all
registrations of DDVP that might cause mammals to be exposed to DDVP
residues should be cancelled.

EFED Response:

The Agency risk assessment includes, as part of the risk
characterization, an evaluation of the potential probability of
individual effects for exposures that may occur at the established
endangered species level of concern.  This probability is calculated
using the established dose/response relationship and the median lethal
dose estimate for the study used to establish the toxicity endpoint for
the endangered taxa.  This approach, which is based on commonly known
and employed basic toxicology dose-response analyses, is explained on
page 64 of the Overview of the Ecological Risk Assessment Process in the
Office of Pesticide Programs (2004) document and was endorsed by the
Fish and Wildlife Service and the National Marine Fisheries Service. 
The Agency uses the Individual Effect Chance spreadsheet tool to perform
these calculations.  The information gained from this analysis is for
descriptive purposes and not for determining if there are adverse
effects.

Comment:

Page 48: The Agency has made the equivalent of “may affect” findings
for endangered species: “The screening level risk assessment has
identified potential concerns for indirect effects on listed species for
those organisms dependent upon freshwater invertebrates, birds, and
mammals. In light of the potential for indirect effects, the next step
for EPA, FWS, and the NMFS is to identify which listed species and
critical habitat are potentially implicated.” The Agency has failed
for far too many years to require adequate data from registrants to
identify endangered species at risk from DDVP. According to EPA, “[a]t
present, the information reviewed by EPA does not permit use of either
analytical approach to make a definitive identification of species that
are potentially impacted indirectly or critical habitats that are
potentially impacted directly by the use of the pesticide.” In light
of this failure, the Agency cannot conclude that continued use of DDVP
will not cause unreasonable adverse effects on the environment under
FIFRA, because it admittedly causes adverse effects among endangered
species. Accordingly, the Agency must cancel all uses that might lead to
exposure of any endangered species.

EFED Response:

See response to NRDC comment immediately below.  

Comment:

Page 49: The Agency has made a “may affect” finding for threatened
or endangered fish species: “Indirect effects to endangered/threatened
fish that depend on freshwater invertebrates as a primary source of
food, as well as larger aquatic animals that rely on aquatic
(freshwater) invertebrate populations as a food source may be affected
by the direct or chronic effects of dichlorvos use.” (emphasis added).
Based on this finding, EPA must immediately commence consultations with
the Fish and Wildlife Service or National Marine Fisheries Service, as
appropriate. EPA’s failure to do so violates the Endangered Species
Act.

EFED Response:

The commentor mischaracterizes EPA’s assessment of potential risks in
the ecological risk assessment as a “may affect” finding.  Although
the risk assessment includes a screening level assessment for threatened
and endangered species, it does not include the species-specific
assessment that EPA uses to make effects determinations.  More
precisely, EPA assesses the potential risks to species listed as
endangered or threatened under the ESA (listed species) as part of its
overall evaluation of the environmental impacts of any given pesticide,
in this case DDVP.  In connection with such a screening level
assessment, EPA requires pesticide companies to conduct studies on a
wide variety of surrogate animals, insects, and other wildlife, and to
submit information that allows EPA to estimate the amount of pesticide
residue that will remain in the environment following its use.  EPA then
reviews and analyzes these data, along with information from the public
literature, to estimate levels of pesticide residues in the environment
that will be unlikely to cause any adverse effects on non-target
species.  These screening level assessments incorporate a number of
conservative assumptions about potential exposure and species
sensitivity so there is little chance an assessment will underestimate
the potential risks to listed species.  These conservative assumptions
may, however, lead to assessments indicating the possibility of a risk
when, in fact, no risk to listed species exists.  Accordingly, the
screening level assessment is not a “may adversely affect”
determination pursuant to the ESA.

If a screening level assessment indicates that a non-target species
could potentially be adversely affected by use of a product, EPA
considers a range of choices, including imposing restrictions on the use
of the pesticide to mitigate possible risks and conducting further
analysis to better characterize the spatial and temporal extent and
magnitude of any risk.  In the case of listed species, further
assessment, on a species-by-species basis, often involves determining
whether a particular listed species (or its protected habitat) will be
in or near areas where a pesticide is or has been used, which is a
substantial, additional effort.  Historically, EPA has conducted
screening level risk assessments for listed species and other non-target
wildlife during reregistration.  These assessments (and the actions
taken by EPA if a screening level assessment suggests there may be risks
of concern) are documented in EPA's FIFRA § 4 Reregistration
Eligibility Determinations (“REDs”).  Given the available resources
and the statutory deadlines for performing reregistration, EPA does not
believe it is feasible routinely to perform follow-on species-specific
assessments for listed species as part of a RED for an individual
pesticide.  Nor do we think it would be prudent public policy to delay
implementation of needed public health and environmental protections
identified through reregistration until we finished performing
species-specific assessments on listed species.  Therefore, as we have
acknowledged, completion of species-by-species assessments leading to
effects determinations, and any consultations necessary under ESA §
7(a)(2), will generally be a post-RED activity.  

Nonetheless, it is important to note that, even in the absence of
completed species-specific effects determinations, EPA has used its
screening level assessments to put in place measures that reduce
exposure and improve the protections for non-target species, including
listed species.  Such measures for DDVP included cancellation of the
crack and crevice use and all uses on lawns, turf, and ornamentals. 

Comment:

Page 50: The risk assessment for aquatic organisms fails to consider
exposure via ingestion: “The screening assessment does not consider
the ingestion pathway. This exposure may occur through ingestion of
contaminated vegetation, invertebrates, or other exposed prey items.”
No explanation is given for EPA’s failure to consider this source of
exposure. In addition, dermal exposure is also left out of the
assessment: “The screening assessment does not consider dermal
exposure. Dermal exposure may occur through one potential source:
contact with contaminated water. The available measured data related to
aquatic wildlife dermal contact with pesticides are extremely
limited.” If the reason for these gaps is inadequate data, the Agency
should have required this data during reregistration. In the absence of
data necessary to reliably characterize these exposure routes, the
Agency cannot determine that uses that might result in exposure to
aquatic organisms will not cause unreasonable adverse effects.
Accordingly, these uses should be cancelled.

EFED Response:

Currently, the Agency’s risk assessment methodologies do not account
for these routes of exposure.  For clarity and transparency, this was
identified in the assumptions and limitations section of the risk
assessment (see page 50).  Nevertheless, the conceptual model assumes
that DDVP use is widespread; on the scale of a major crop use
agrochemical.  Estimated environmental concentrations are likely very
conservative given the standard, underlying assumptions built into the
exposure models.

Comment:

Page 51: EPA acknowledges that the LD50/sq. ft. calculation is not
adequately protective:  “It does not address subacute behavioral or
physiological effects that may occur prior to mortality and yet can
still have a profound sub-lethal effects [sic] on an organisms [sic]
ability to survive and reproduce. As such, this index may underestimate
ecological hazard from sub-lethal exposures.” Based on this finding
alone, it is clear that the Agency cannot justifiably conclude that the
exposure of wildlife to DDVP will not cause unreasonable adverse
effects. Accordingly, all uses that may result in exposure to wildlife
should be cancelled. (Furthermore, the Agency must be mindful that for
endangered species, even a single LD50 exposure to an endangered specie
is unlawful, regardless of the size area it occurs in).

EFED Response:

The Agency’s Overview of the Ecological Risk Assessment Process in the
Office of Pesticide Programs (2004) document, which was endorsed by the
Fish and Wildlife Service and National Marine Fisheries Service,
describes the assumptions and limitations regarding sublethal effects. 
For an acute risk assessment, the screening risk assessment relies on
the acute mortality endpoint as well as a suite of sublethal responses
to the pesticide, as determined by the testing of species response to
chronic exposure conditions and subsequent chronic risk assessment. 
Examples of these sublethal endpoints for birds and mammals include the
following:

	Birds:		Reproduction		Maternal weight

						Eggs laid/hen

						Eggs cracked

						Eggshell thickness

						Viable embryos

						Hatchling number

						14-day survivors

						Gross necropsy

						Observations of other clinical signs

Mammals:	Two-Generation	Total panel of reproduction parameters

		Reproduction		including:

					Histopathology, parental and offspring

					growth, weight, mating, lactation,

					gonadal development milestones,

					sexual organ performance and offspring

					production

A risk assessment team has the option of considering additional
sublethal data in the assessment.  This option is exercised on a
case-by-case basis and only after careful consideration of the nature of
the sublethal effect measured and the extent and quality of available
data to support establishing a plausible relationship between the
measure of effect (sublethal endpoint) and the assessment endpoints. 
This option includes a determination of whether there are clear,
reasonable, and plausible links between the sublethal effect and
survival or reproductive capacity of organisms in the field in
accordance with the screening assessment endpoints of survival and
reproduction capacity.  The Agency documents the findings of such
evaluations of additional sublethal effects in the effects assessment
and includes a discussion of their potential effects upon the confidence
of the overall risk assessment conclusions.  The Agency anticipates
that, through the SAP and related external peer-review processes in the
scientific community, accepted risk assessment practices will continue
to advance in this area.

Comment:

Pages 52-53: The Agency provides a detailed list of limitations of its
ecological risk assessment. It is commendable that these limitations are
frankly identified. However, the Agency fails to recognize that these
uncertainties, almost all of which would tend to result in
understatement of the risks, prevent EPA from finding that any uses of
DDVP which might lead to exposure of wildlife – and endangered species
in particular – will not cause unreasonable adverse effects on the
environment. Accordingly, all such uses should be cancelled.

EFED Response:

The DDVP screening level risk assessment uses the best available
science.  As pointed out in the Overview of the Ecological Risk
Assessment Process in the Office of Pesticide Programs (2004) document,
which was endorsed by the Fish and Wildlife Service and the National
Marine Fisheries Service, there will always be uncertainties in
screening level ecological risk assessments.  As was done in the DDVP
assessment, it is important to describe the assumptions and limitations
so that they may be taken into account in the risk management decision.

Respectfully submitted,

/s/ Aaron Colangelo

_____________________________

Aaron Colangelo

Natural Resources Defense Council

1200 New York Ave NW, Suite 400

Washington DC 20005

cc: Dayton Eckerson, SRRD

Response to Comments Received 8/8/05 from Robert Crandall, Cereal Food
Processors, Inc., 2001 Shawnee Mission Parkway, Mission Woods, KS 66205

Subject: OPP-2002-0302 (Dichlorvos (DDVP) Revised Ecological Risk
Assessment; Notice of Availability)

August 8, 2005

Comment:

Our company has flour mills located in several states across the United
States. Each location fogs regulary with DDVP during months they do not
perform a general fumigation. Some plants do one general fumigation a
year while others do two general fumigations a year. From this you can
tell we depend very heavily on the proven effectiveness of DDVP.

DDVP was reformulated a few years ago with a different carrier. As a
result one of our plants had to use a different material to fog with
while the new formulation was going through approval for their state.

The plant had to increase the amount of residual pesticides applied
inside the plant. This not only took more time and money but was no
where near as effective as fogging with DDVP. Before the plant had to
use the alternate product they had reduced the application of residual
pesticides used inside the plant to almost nothing. Nothing being once a
year as opposed to monthly.

After the new formulation was approved the monthly fogging with DDVP
allowed the plant to cut back on the use of residual pesticides to the
levels previous to the reformulation.

For this reason we are very much in favor of the re-registration of
Dichlorvos (DDVP).

Robert Crandall

Manager of Food Safety & Sanitation

Cereal Food Processors, Inc.

Cc: 	John Erker, CFP

	Royal Denning, CFP

EFED Response:

These comments appear to be concerning DDVP use inside flour mills. 
Therefore, this use would not be included in the ecological risk
assessment and requires no EFED response.

 PAGE   

 PAGE   1 

		

